View in EDS HTML Full Text PDF Full Text

The Cognitive Diagnostic Model Analysis of the Relationship between Verbal and Quantitative Skills with Background Characteristics: Evidence from the Swedish Scholastic Assessment Test 2023

Saved in:
Bibliographic Details
Title: The Cognitive Diagnostic Model Analysis of the Relationship between Verbal and Quantitative Skills with Background Characteristics: Evidence from the Swedish Scholastic Assessment Test 2023
Language: English
Authors: Könül Karimova (ORCID 0000-0003-4840-3573), Inga Laukaityte (ORCID 0000-0001-7282-5384), Christina Wikström (ORCID 0000-0002-4625-4853)
Source: Journal of Advanced Academics. 2026 37(2):364-392.
Availability: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
Peer Reviewed: Y
Page Count: 29
Publication Date: 2026
Document Type: Journal Articles
Reports - Research
Education Level: High Schools
Secondary Education
Higher Education
Postsecondary Education
Descriptors: Foreign Countries, Aptitude Tests, High Stakes Tests, Verbal Ability, Mathematics Skills, Cognitive Tests, Models, Test Wiseness, Student Characteristics, Measurement, Gender Differences, Educational Attainment, Age Differences, Student Evaluation, Item Response Theory, Multiple Choice Tests, Reading Tests, Reading Comprehension, College Students
Geographic Terms: Sweden
DOI: 10.1177/1932202X251371339
ISSN: 1932-202X
2162-9536
Abstract: For decades, researchers have sought analytic methods that yield diagnostic information about test-takers while ensuring that high-stakes tests remain free of construct-irrelevant bias. This study applies a general cognitive-diagnostic-model (CDM) framework to analyze the quantitative and verbal skills (QVS) assessed by the Swedish Scholastic Aptitude Test (SweSAT). A three-step latent-class logistic-regression approach was used to investigate the relationship between test-takers' background characteristics and performance in each domain subskill. The analysis was conducted using representative data from 41,451 test-takers from the 2023 administration of the SweSAT, focusing on performance across four quantitative and four verbal subskills, and examining test characteristics. The results showed that the CDM method was appropriate for analyzing QVS, with evidence of measurement invariance across sex, age, and educational levels in the subskills. Additionally, the findings revealed distinct associations between sex, educational level, and age with performance in each QVS subskill. Implications for equitable selection in higher education are discussed.
Abstractor: As Provided
Entry Date: 2026
Accession Number: EJ1504546
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFb_qBMbDiB9ICJH2xQYLAIAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDD7ZZ8KUlUVJYoVbswIBEICBmydzfcabNyIH5qxT1Bril2iY5ZOYfwBJzufKwKiCIT4rHgO0NUTxiPN1kL2xFT-3UyLZ1u18mVWKKgcmTg2QrnJVIlxVAfKGSx3zxHFJnMrcsMWTLieWIELxlNytxmztkIehhudPUCJwPW6WvHGq-GMRCgjLM2cpsuuaehIL-4LoZP3fSJ2Vuk4TRwIj4q4jKDcJGltxT1PFyBD4
Text:
  Availability: 1
  Value: <anid>AN0193364109;[261p]01may.26;2026May04.05:11;v2.2.500</anid> <title id="AN0193364109-1">The Cognitive Diagnostic Model Analysis of the Relationship Between Verbal and Quantitative Skills With Background Characteristics: Evidence From the Swedish Scholastic Assessment Test 2023 </title> <p>For decades, researchers have sought analytic methods that yield diagnostic information about test-takers while ensuring that high-stakes tests remain free of construct-irrelevant bias. This study applies a general cognitive-diagnostic-model (CDM) framework to analyze the quantitative and verbal skills (QVS) assessed by the Swedish Scholastic Aptitude Test (SweSAT). A three-step latent-class logistic-regression approach was used to investigate the relationship between test-takers' background characteristics and performance in each domain subskill. The analysis was conducted using representative data from 41,451 test-takers from the 2023 administration of the SweSAT, focusing on performance across four quantitative and four verbal subskills, and examining test characteristics. The results showed that the CDM method was appropriate for analyzing QVS, with evidence of measurement invariance across sex, age, and educational levels in the subskills. Additionally, the findings revealed distinct associations between sex, educational level, and age with performance in each QVS subskill. Implications for equitable selection in higher education are discussed.</p> <p>Keywords: quantitative and verbal skills; background characteristics; cognitive diagnosis models; three-step latent analysis; differential item functioning; logistic regression; Swedish Scholastic Aptitude Test</p> <p>Throughout the decades, researchers used various analysis methods to study quantitative and verbal skills (QVS; e.g., [<reflink idref="bib2" id="ref1">2</reflink>]; [<reflink idref="bib14" id="ref2">14</reflink>]; [<reflink idref="bib31" id="ref3">31</reflink>]; [<reflink idref="bib35" id="ref4">35</reflink>]; [<reflink idref="bib48" id="ref5">48</reflink>]). QVS were considered crucial for further improvement of other academic skills (e.g., [<reflink idref="bib21" id="ref6">21</reflink>]). Many large-scale assessments (LSA), both high- and low-stakes, aim to measure students' QVS across various age groups. The Scholastic Aptitude Test (SAT) is a widely recognized example of a high-stakes LSA designed to assess the QVS necessary for academic success in college. By analyzing SAT performance data, researchers can gain a deeper understanding of how these two areas of competence—verbal and quantitative—interact and are influenced by background characteristics such as socioeconomic status (SES), sex, race, and educational environment. Detailed diagnostic information about test-takers underpins enrichment programs that equitably support high-ability students regardless of background characteristics and promote their full potential development. By extending beyond traditional curricula and highlighting systemic inequities, such programs demand rigorous analytic methods to relate QVS to background characteristics. Applying the appropriate method can yield detailed information about test-takers, providing insights into facilitating equal access to higher learning opportunities and/or addressing the underlying factors that lead to gaps in opportunity, excellence, and equity.</p> <p>Some studies have examined the validity of the SAT for predicting scores (e.g., [<reflink idref="bib31" id="ref7">31</reflink>]), while others have explored associations between background variables and ability profiles (e.g., [<reflink idref="bib35" id="ref8">35</reflink>]). However, research in this area has generally adopted a broad perspective and has not directly assessed how QVS may differ across subgroups defined by sex, age, and educational level. [<reflink idref="bib35" id="ref9">35</reflink>] applied an empirical clustering method to identify profiles of predictor variables that best predicted college student outcomes. They identified five distinct student groups, each differing based on background information. The researchers conducted studies to examine the relationship between these student subgroups' mastery of QVS and their respective background characteristics. Consequently, when aggregate score differences are observed—for example, higher mean quantitative scores for men or higher mean vocabulary scores for older adults—these disparities should not be conflated with bias unless item-level analyses demonstrate differential item functioning (DIF; [<reflink idref="bib30" id="ref10">30</reflink>]).</p> <p>Therefore, this study makes a clear distinction between performance differences attributable to unequal opportunities to learn or practice and true measurement bias that compromises the validity of test scores. Furthermore, most studies have only focused on the relationship between grade point average and background characteristics, neglecting domain multidimensionality within QVS ([<reflink idref="bib2" id="ref11">2</reflink>]; [<reflink idref="bib14" id="ref12">14</reflink>]; [<reflink idref="bib35" id="ref13">35</reflink>]; [<reflink idref="bib36" id="ref14">36</reflink>]). For instance, [<reflink idref="bib33" id="ref15">33</reflink>] ([<reflink idref="bib23" id="ref16">23</reflink>]; [<reflink idref="bib47" id="ref17">47</reflink>]), Program for International Student Assessment (PISA) 2012 ([<reflink idref="bib46" id="ref18">46</reflink>]), and Progress in International Reading Literacy Study (PIRLS) 2016 ([<reflink idref="bib40" id="ref19">40</reflink>]) applied cognitive diagnosis models (CDMs) for students to examine their mastery of reading and mathematics skills. However, these studies have not examined how subgroups differ in the mastery of QVS based on background information. Substantially, [<reflink idref="bib25" id="ref20">25</reflink>] investigated the mastery profiles of digital skills of students applying general CDM and found measurement invariance across gender and SES. However, the limitation of [<reflink idref="bib25" id="ref21">25</reflink>] study was that they did not investigate the age variation, revealing that applying multiple-group CDMs ([<reflink idref="bib27" id="ref22">27</reflink>]) in the examination of students from the three cohorts could contribute to CDM application research.</p> <p>Thus, the present study examines how background information is related to QVS performance at the skill level. This research used a three-step approach under a generalized CDM framework to identify test-takers' profiles on the QVS and investigate the test-takers' background information that could be related to their performance on the QVS.</p> <hd id="AN0193364109-2">Swedish Scholastic Aptitude Test (SweSAT)</hd> <p>The SweSAT is a large-scale high-stakes test, developed for the assessment of the test-takers' general ability for studies and applied to the selection process for postsecondary education. The first SweSAT administration was conducted in 1977. Until the spring of 1991, only a limited group of applicants based on age and work experience were eligible to enroll in the SweSAT assessment. Since the autumn of 1991, anyone could take the SweSAT, but there is now a lower age limit of 18 years ([<reflink idref="bib36" id="ref23">36</reflink>]). Although the current version of the SweSAT has been used since 2011, the test structure and subtest descriptions are still best documented in [<reflink idref="bib36" id="ref24">36</reflink>], as no updated reference manual has been published since. The validity of test scores is for 8 years, and the test-taker is allowed to enroll in several test sessions. The SweSAT is composed of seven subtests that assess verbal and quantitative skills, the ability to utilize information, and overall knowledge. The verbal section consists of subsets vocabulary (WORD), Swedish reading comprehension (READ), sentence completion (SEC), and English reading comprehension (ERC) and the quantitative section consists of subsets mathematical problem-solving (XYZ), quantitative comparisons (QCs), data sufficiency (DS), and diagrams, tables, and maps (DTM).</p> <hd id="AN0193364109-3">Verbal Section</hd> <p>The purpose of the WORD subtest is to evaluate comprehension of vocabulary and concepts ([<reflink idref="bib36" id="ref25">36</reflink>]). The WORD subtest encompasses both Swedish terms and loanwords, particularly those anticipated to appear in scholarly texts. The selected texts of the READ subtest encompass a variety of topics and linguistic styles, primarily focusing on popular science, opinion articles, reviews, and similar genres. These texts are sourced from contemporary magazines, newspapers, and books, reflecting the types of written materials that university students are likely to encounter in their academic pursuits. The ERC items are designed to identify key information within the text rather than focusing on trivial details, emphasizing central arguments and conclusions. Each ERC test includes authentic texts covering a wide range of topics, including the arts, social sciences, technology, and natural sciences. The purpose of the SEC subtest is to assess the students' ability to articulate words and phrases within a given context. This subtest has been presented since 2011.</p> <hd id="AN0193364109-4">Quantitative Section</hd> <p>The XYZ explores the test-taker's ability to address mathematical challenges across various domains, including arithmetic, algebra, geometry, function theory, and statistics ([<reflink idref="bib36" id="ref26">36</reflink>]). Each parallel test comprised 12 multiple-choice items, which test-takers are expected to respond to within 12 minutes. The content of items consisted of algebra and geometry. The QC explores the test-taker's ability to conduct QCs across various domains, including arithmetic, algebra, geometry, function theory, and statistics. Each parallel test consisted of 10 multiple-choice items, which test-takers are expected to respond to within 10 minutes. The content of items consisted of comparisons within algebra or geometry. The DS subtest test-takers are tasked with determining whether sufficient information is available to resolve a specific problem. The format of the items is standardized, meaning that the answer choices remain consistent across all test items. The DTM subset aims to evaluate the capacity to interpret data presented in diagrams, tables, maps, and other graphical formats. Educational resources frequently include such graphical representations from which students must extract pertinent information. Each figure set includes two questions, each offering five response options.</p> <hd id="AN0193364109-5">The Relationship Between Background Information and QVS</hd> <p>Most studies have examined the background characteristics related to the QVS of SAT and found a relationship between background variables such as socioeconomic factors with QVS of SAT scores (e.g., [<reflink idref="bib48" id="ref27">48</reflink>]) and external variables such as self-rated performance, class absenteeism, organizational citizenship behavior, intent to quit university, satisfaction with SAT, and American College Test scores (e.g., [<reflink idref="bib35" id="ref28">35</reflink>]). [<reflink idref="bib13" id="ref29">13</reflink>] examined gender gap variation across assessment types, and they revealed that grades explained much of the variation in the gender gap, indicating that females outperformed males in school grades. They showed the possibility of the SAT to penalize women. [<reflink idref="bib45" id="ref30">45</reflink>] studied gender-related DIF and found that several vocabulary items favored women, while several sentence completion item formats favored men. [<reflink idref="bib36" id="ref31">36</reflink>] examined the relationship between SweSAT test-takers' educational level with their results and revealed that a higher level of education led to better results on the test. Further, they investigated the association of various age groups with QVS, and they also explored a positive relationship between vocabulary and age, meaning that the higher the age, the better the test-takers' performance on the test. In contrast, DS and DTM showed a weak negative relationship with age, yielding that the higher the age the lower the performance. However, [<reflink idref="bib36" id="ref32">36</reflink>] used the traditional item analysis method, such as item response theory (IRT) in their study without the opportunity to apply [<reflink idref="bib41" id="ref33">41</reflink>] three-step latent approach to prevent biased coefficient estimation. [<reflink idref="bib24" id="ref34">24</reflink>] applied one of the CDMs such as deterministic, inputs, noisy, "and" gate model (DINA) to examine the proportions of correct answering and mastering mathematics and science skills required to solve an item by using TIMSS 2007 data. They revealed that the application of CDM provided fine-grained information that can be transferred directly to classroom application at the attribute level. [<reflink idref="bib40" id="ref35">40</reflink>] examined the 2016 PIRLS data, using other CDMs, such as the log-linear cognitive diagnosis modeling (LCDM), and he found that this methodology also provided more fine-grained diagnostic information about test-takers' reading skills than traditional test scoring. [<reflink idref="bib46" id="ref36">46</reflink>] formed three dimensions from 11 attributes based on the mathematical framework of PISA 2012 by applying the linear logistic model (LLM), and they explored the variation in the test-takers' mastery of the 11 cognitive attributes from three dimensions of content, process and context across 10 countries.</p> <hd id="AN0193364109-6">Statistical Methods for Examination of the Relationship Between QVS and Background Informatio...</hd> <p>Multiple statistical methods have been applied to examine the relationship between background information and the QVS of LSA. Some research used simple statistical methods, such as correlation matrixes and percentage of variances ([<reflink idref="bib33" id="ref37">33</reflink>]; [<reflink idref="bib48" id="ref38">48</reflink>]), while other studies used regression or multilevel logistic regression analysis to explore variable predictability of QVS scores ([<reflink idref="bib2" id="ref39">2</reflink>]; [<reflink idref="bib14" id="ref40">14</reflink>]; [<reflink idref="bib31" id="ref41">31</reflink>]). Thus, these studies applied conventional methods such as classical test theory (CTT) or IRT to provide the psychometric properties of measurement they employed. This study applies CDM to examine the test-takers' performance of QVS, and a three-step latent logistic regression approach ([<reflink idref="bib18" id="ref42">18</reflink>]; [<reflink idref="bib41" id="ref43">41</reflink>]) is used to analyze the variation in the relationship between background characteristics and test-takers' mastery status for each skill. Mastery status is considered as binary (0/1).</p> <hd id="AN0193364109-7">Cognitive Diagnostic Models</hd> <p>Cognitive diagnostic models or classification models are discrete latent class models that can provide detailed information about the test-takers' weaknesses and strengths at the attribute level (e.g., [<reflink idref="bib4" id="ref44">4</reflink>]; [<reflink idref="bib8" id="ref45">8</reflink>]; [<reflink idref="bib15" id="ref46">15</reflink>]; [<reflink idref="bib20" id="ref47">20</reflink>]; [<reflink idref="bib32" id="ref48">32</reflink>]; [<reflink idref="bib42" id="ref49">42</reflink>]; [<reflink idref="bib43" id="ref50">43</reflink>]). In contrast to traditional IRT and CTT models, CDMs can shed light on test-takers' classification corresponding to a skill, contrary to the cumulative nature of CTT and IRT models ([<reflink idref="bib9" id="ref51">9</reflink>]). Similar score test-takers can vary in mastery of various skill attributes. Therefore, to obtain detailed information on the test-takers' true performance, this study applied CDMs to examine the test-takers' performance level of QVS.</p> <p>Being a type of confirmatory modeling method, CDM consists of two parts: the measurement part and the latent part. The measurement part depicts how observed responses explain latent attributes established earlier by the Q-matrix ([<reflink idref="bib38" id="ref52">38</reflink>]). The Q-matrix is a binary matrix in which 1s denote the presence of matching characteristics needed to properly answer a particular question and 0s the absence of such attributes. The number of latent classes in the present study is determined by the number of assessed qualities; each part of SweSAT, either quantitative or verbal, consists of four skills which yield 2<sups>4</sups> = 16 latent classes. Generally, but not always, a binary matrix containing each person's replies (rows) to the test items (columns) serves as the input for CDMs. Test-takers are categorized into a latent class based on their performance on each item following computation. CDMs are classified into two types: specific and general. Specific CDMs can be conjunctive or disjunctive, depending on various diagnostic goals or condensation rules ([<reflink idref="bib29" id="ref53">29</reflink>]). Being a conjunctive model, deterministic inputs, noisy "and" gate (DINA; e.g., [<reflink idref="bib7" id="ref54">7</reflink>]) model entails a correct item response from all the required attributes contrasting to deterministic inputs, noisy "or" gate (DINO; [<reflink idref="bib39" id="ref55">39</reflink>]) model which is a disjunctive model, entailing a correct item response from at least one required attribute that test-taker has mastered. Furthermore, [<reflink idref="bib8" id="ref56">8</reflink>] contributed to CDM research by introducing additive CDM (A-CDM) as well as [<reflink idref="bib29" id="ref57">29</reflink>] presented a LLM entailing that an increase of success on an item related to the performance of each required attribute which is independent of the contributions of the other attributes. Recent studies applied the reparametrized unified model (i.e., the fusion model) or reduced reparametrized unified model (i.e., R-RUM). The difference between RUM and R-RUM is in the inclusion of residual parameters that indicate the magnitude of the item response function based on skills other than those assigned by the Q-matrix ([<reflink idref="bib3" id="ref58">3</reflink>]). R-RUM does not incorporate this parameter.</p> <p>Further, being an example of general CDM, the generalized deterministic input, noisy "and" gate (G-DINA; [<reflink idref="bib8" id="ref59">8</reflink>]) model has been proposed to facilitate the assumptions of the DINA model and to examine various success probabilities for an item based on various subsets of attributes. The saturated G-DINA model considers all possible main and interaction effects based on required attributes. The reduced models involve the DINA model, the DINO model, the A-CDM, the LLM, and the R-RUM. Since the G-DINA model is considered a general CDM framework and provides detailed information about item effects at the item level, the present study applied the G-DINA model framework to analyze the test-takers' performance and the relationship of performance with background information.</p> <hd id="AN0193364109-8">Measurement Invariance</hd> <p>Since test fairness is a critical concern in any assessment, determining whether DIF is present or absent is essential. Therefore, examining measurement invariance is necessary (e.g., American Educational Research Association [AERA], American Psychological Association [APA], & National Council on Measurement in Education [NCME], 2014; [<reflink idref="bib27" id="ref60">27</reflink>]). If test-takers from various groups with similar abilities have an equal chance of responding to the items correctly, the item is considered a non-DIF item ([<reflink idref="bib28" id="ref61">28</reflink>]). Similarly, under the CDM framework, if test-takers from various groups with similar mastery profiles have different chances to respond to an item, that item is considered a DIF item ([<reflink idref="bib16" id="ref62">16</reflink>]). Since the existence of DIF items can indicate test validity problems, the present study needs to ascertain that measurement sets equal opportunities for each group of test-takers in high-stakes LSA tests. Thus, this study examines the DIF of QVS items on the SweSAT across various background characteristics (sex, age, and educational level).</p> <hd id="AN0193364109-9">Aims of This Study</hd> <p>Most studies have evaluated the test takers' performance from a general perspective neglecting differentiation at the attribute level of each skill in LSA. Furthermore, there is a lack of studies that explored the relationship between the test takers' background information and the mastery of QVS. Therefore, this study investigates the test takers' mastery of QVS and the relationship between background characteristics and performance level in the high-stakes large-scale assessment, SweSAT, by applying a CDM framework and integrating a three-step latent class analysis. In this study, the following research questions are addressed:</p> <p></p> <ulist> <item> To what extent can the application of CDMs provide differentiation within QVS achievement in LSA such as SweSAT?</item> <p></p> <item> What are the mastery profiles of SweSAT test takers' QVS defined by CDMs?</item> <p></p> <item> Does QVS hold measurement invariance across test takers' background characteristics such as sex, age, and educational level?</item> <p></p> <item> How are test takers' background characteristics related to their performance of QVS?</item> </ulist> <hd id="AN0193364109-10">Method</hd> <p></p> <hd id="AN0193364109-11">Sample and Data</hd> <p>The participants of this study were 41,451 test takers of the SweSAT 2023, autumn administration. Among them, 54.1% were women. Most test-takers, 68.5%, were up to 20 years old, and 81.7% of all participants had completed or were at upper secondary school, lasting 3 or more years (see Table 1).</p> <p>Table 1. Descriptive Statistics of Background Information of the SweSAT2023 Participants.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="char" char="." /></colgroup><thead><tr><th align="left">Covariate</th><th align="left">Category</th><th align="left"><italic>N</italic></th><th align="left">Percentage (%)</th></tr></thead><tbody><tr><td>Sex</td><td>Male</td><td>19,041</td><td>45.9</td></tr><tr><td /><td>Female</td><td>22,410</td><td>54.1</td></tr><tr><td>Age</td><td>Up to 20 years old</td><td>28,378</td><td>68.5</td></tr><tr><td /><td>21–24 years old</td><td>7,984</td><td>19.3</td></tr><tr><td /><td>25–29 years old</td><td>2,412</td><td>5.8</td></tr><tr><td /><td>30–39 years old</td><td>1,828</td><td>4.4</td></tr><tr><td /><td>40 years old or above</td><td>849</td><td>2.0</td></tr><tr><td>Educational level</td><td>Compulsory school</td><td>785</td><td>1.9</td></tr><tr><td /><td>Folk high school</td><td>296</td><td>0.7</td></tr><tr><td /><td>Upper secondary school, less or 2 years</td><td>1,197</td><td>2.9</td></tr><tr><td /><td>Upper secondary school, 3 or more years</td><td>33,850</td><td>81.7</td></tr><tr><td /><td>Postsecondary education, less than 3 years</td><td>2,925</td><td>7.1</td></tr><tr><td /><td>Postsecondary education, 3 or more years</td><td>1,585</td><td>3.8</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note:</emph> The term "sex" is used in this study rather than "gender" because the SweSAT dataset only included binary options (male/female). This terminology reflects the structure of the collected data, acknowledging that gender is a broader and socially constructed category not captured by the test instrument.</p> <p>The SweSAT consists of 160 dichotomously scored multiple-choice items testing two main skills: quantitative and verbal. Each test taker completes two quantitative subtests (40 items each) and two verbal subtests (40 items each). One quantitative subtest (40 items) is designed to assess four quantitative dimensions of QVS: mathematical problem-solving (A1, 12 items), quantitative comparison (A2, 10 items), DS (A3, 6 items), and DTM(A4, 12 items). Similarly, each verbal section is designed to assess four verbal dimensions of QVS: vocabulary (A1, 10 items), Swedish reading comprehension (A2, 10 items), sentence completion (A3, 10 items), and ERC (A4, 10 items).</p> <p>The design of SweSAT subtests corresponded to the Q-matrix developed by content experts ([<reflink idref="bib36" id="ref63">36</reflink>]). Before administration, each item of the SweSAT is pre-tested and reviewed by various domain experts, such as subject matters. Before inclusion in the final test, each item has been approved by up to ten individuals. Detailed information about SweSAT can be obtained from [<reflink idref="bib36" id="ref64">36</reflink>] study. Using the design structure of SweSAT data, this study conducts a confirmatory CDM framework for the Q-matrix validation of the SweSAT design structure. The Q-matrix of QVS is provided in Appendix A. Data is available at Open Science Framework (https://osf.io/hf8cz/).</p> <hd id="AN0193364109-12">Background Information</hd> <p>This study analyzes the test takers' performance and examines background information based on educational level, age, and sex. Unlike other studies, this analysis is based on the CDM framework using a three-step latent class approach. Specifically, the G-DINA model framework was applied to explore test-takers' mastery profiles of QVS in high-stakes tests and to examine the relationship between the test-takers' background information and mastery profile of QVS through logistic regression conducted via the three-step latent analysis. This approach provides detailed information about how the test-takers' background characteristics are related to mastery of each attribute. Educational level, age, and sex were the three categorical covariates used in the analysis (see Table 1). The educational level included six levels, with the sixth level representing the highest level of education (postsecondary education, 3 or more years). Age included five levels and sex included two levels. Additionally, a DIF analysis was conducted to examine the presence of between-group differences based on background characteristics.</p> <hd id="AN0193364109-13">Statistical Analysis</hd> <p></p> <hd id="AN0193364109-14">Cognitive Diagnostic Modeling Analysis</hd> <p>One subtest of the QVS, containing 40 items, assesses a single quantitative skill, while another subtest, also with 40 items, assesses a single verbal skill. Both are based on the Q-matrix. To make interpretations easier and improve classification rates, besides the saturated G-DINA model, most researchers recommend using reduced models (e.g., [<reflink idref="bib22" id="ref65">22</reflink>]). Thus, following [<reflink idref="bib25" id="ref66">25</reflink>] study, CDM analysis consisted of two parts. The saturated G-DINA model was conducted in the first part to fit the data. The second part involved determining the proper CDM (e.g., DINA, DINO, A-CDM, LLM, and R-RUM) for each skill by conducting the Wald test with a.05 significance level. Following [<reflink idref="bib10" id="ref67">10</reflink>] CDM analysis procedures, the G-DINA model was deemed suitable for a particular item when none of the reduced models were selected. However, the reduced model with the highest <emph>p</emph>-value was chosen when many appropriate reduced models existed. Subsequently, to measure the item characteristics and test-takers' mastery profiles, the relevant CDMs based on the Wald test were applied to fit the data. Thus, following recommendations, this study used absolute fit measures such as the residual between observed and predicted Fisher-transformed correlation (<emph>r</emph>) and log-odds ratios of item pairs (<emph>l</emph>) ([<reflink idref="bib4" id="ref68">4</reflink>]). Further, we adopted Bayesian Information Criterion (BIC) as the primary relative-fit statistic because our large sample size (<emph>n</emph> = 41,451) inflates [<reflink idref="bib1" id="ref69">1</reflink>] information criterion penalty only modestly ([<reflink idref="bib19" id="ref70">19</reflink>]). Following Jeffreys, the BIC values above 10 indicate decisive evidence for the superior model; values below 2 suggest that competing models fit the data equally well.</p> <hd id="AN0193364109-15">Differential Item Functioning (DIF)</hd> <p>DIF analysis was performed using the generalized logistic regression-based Wald test to detect whether items functioned differently across test-takers' background characteristics (sex, age, and educational level), independent of their overall ability levels (e.g., [<reflink idref="bib37" id="ref71">37</reflink>]). This method, rather than relying on raw mean differences, allows for the identification of potential measurement bias. Descriptive statistics for the covariates are presented in Table 1.</p> <hd id="AN0193364109-16">Three-Step Latent Logistic Regression Analysis</hd> <p>Following [<reflink idref="bib10" id="ref72">10</reflink>] suggestions, this study integrated covariates into CDMs to examine the correlation between test-takers' background information and their performance of QVS. To avoid bias in the coefficient estimation of logistic regression which can impact the validity of results, a three-step latent class analysis was conducted by applying correction weights for skill-level regression ([<reflink idref="bib18" id="ref73">18</reflink>]; [<reflink idref="bib41" id="ref74">41</reflink>]).</p> <p>This study examined how test-takers' performance on each skill was related to covariates using a three-step latent logistic regression analysis ([<reflink idref="bib18" id="ref75">18</reflink>]). In the first step, CDM was fitted to the QVS data. In the second step, test-takers were assigned to latent classes. In the third step, latent logistic regression was fitted for each QVS. All covariates were dummy-coded. This study conducted two common techniques of regression analysis such as simple logistic regression where each QVS skill regressed on each covariate separately and multiple logistic regression where each QVS skill regressed on all covariates concurrently. Further, effect sizes, such as Cohen's <emph>d</emph> ([<reflink idref="bib6" id="ref76">6</reflink>]), were estimated to evaluate the magnitude of the relationships between test-takers' performance and covariates, with <emph>d</emph> values of.2,.5, and.8 corresponding to small, medium, and large effects, respectively. This study analyzed the SweSAT data using R software with the G-DINA package ([<reflink idref="bib26" id="ref77">26</reflink>]).</p> <hd id="AN0193364109-17">Results</hd> <p></p> <hd id="AN0193364109-18">Cognitive Diagnosis Modeling</hd> <p>The 80 items of the SweSAT-assessed QVS. All 80 items were fitted by applying CDMs in Table 2.</p> <p>Table 2. Selected CDMs for Quantitative and Verbal Skill Items.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Selected model</th><th align="left">Item</th></tr></thead><tbody><tr><td>Quantitative skill</td><td /></tr><tr><td> G-DINA</td><td>1, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40</td></tr><tr><td> LLM</td><td>9</td></tr><tr><td> A-CDM</td><td>2</td></tr><tr><td> R-RUM</td><td>22</td></tr><tr><td>Verbal skill</td><td /></tr><tr><td> G-DINA</td><td>1, 2, 3, 4, 6, 7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40</td></tr><tr><td> LLM</td><td>5, 23</td></tr><tr><td> R-RUM</td><td>8</td></tr></tbody></table> </ephtml> </p> <p>2 <emph>Note:</emph> G-DINA: generalized deterministic inputs, noisy and gate model; LLM: linear logistic model; A-CDM: additive cognitive diagnostic model; R-RUM: reduced reparametrized unified model.</p> <p>Thirty-seven items in both QVS were fitted by the G-DINA model showing that QVS evaluated in these items have all possible interaction and main effects. Since items 5 and 23 in verbal skill and item 9 in quantitative skill showed an additive effect, the LLM, as an additive model under the logit link, was selected to evaluate these items. The A-CDM was appropriate for item 2 in quantitative skill to measure all interaction effects when constrained to zero. The reduced model for item 8 was the R-RUM, which did not include a parameter providing information about the magnitude of a lack of substantial skill.</p> <p>To examine the viability of CDM models in differentiating QVS achievement in high-stakes LSA tests, this study compared the relative and absolute fit statistics of the G-DINA model with those of other models. The relative fit statistics of the six CDA models were estimated, indicating that all other models fit the data in comparison to the G-DINA model (Table 3).</p> <p>Table 3. Model-Data Fit Statistics.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Model</th><th align="left">BIC</th><th align="left"><italic>Npar</italic></th><th align="left">Max. <italic>z</italic>(<italic>r</italic>)</th><th align="left">Max. <italic>z</italic>(<italic>l</italic>)</th></tr></thead><tbody><tr><td>Quantitative skill</td><td /><td /><td /><td /></tr><tr><td> G-DINA model</td><td>1,966,744.88</td><td>95</td><td>59.42(.00)</td><td>51.92(.00)</td></tr><tr><td> DINA</td><td>1,966,743.99</td><td>95</td><td>59.42(.00)</td><td>51.92(.00)</td></tr><tr><td> DINO</td><td>1,966,743.99</td><td>95</td><td>59.42(.00)</td><td>51.92(.00)</td></tr><tr><td> A-CDM</td><td>1,966,743.99</td><td>95</td><td>59.40(.00)</td><td>51.90(.00)</td></tr><tr><td> LLM</td><td>1,966,744.05</td><td>95</td><td>59.41(.00)</td><td>51.92(.00)</td></tr><tr><td> R-RUM</td><td>1,966,744.05</td><td>95</td><td>59.38(.00)</td><td>51.90(.00)</td></tr><tr><td>Verbal skill</td><td /><td /><td /><td /></tr><tr><td> G-DINA</td><td>1,976,090.85</td><td>95</td><td>24.99(.00)</td><td>24.51(.00)</td></tr><tr><td> DINA</td><td>1,976,090.85</td><td>95</td><td>59.42(.00)</td><td>51.92(.00)</td></tr><tr><td> DINO</td><td>1,976,090.85</td><td>95</td><td>24.99(.00)</td><td>24.51(.00)</td></tr><tr><td> A-CDM</td><td>1,976,090.91</td><td>95</td><td>24.95(.00)</td><td>24.48(.00)</td></tr><tr><td> LLM</td><td>1,976,090.62</td><td>95</td><td>25.00(.00)</td><td>24.52(.00)</td></tr><tr><td> R-RUM</td><td>1,976,088.06</td><td>95</td><td>24.94(.00)</td><td>24.47(.00)</td></tr></tbody></table> </ephtml> </p> <p>3 <emph>Note:</emph> G-DINA: generalized deterministic inputs, noisy and gate model; DINO: deterministic inputs, noisy "or" gate; LLM: linear logistic model; A-CDM: additive cognitive diagnostic model; R-RUM: reduced reparametrized unified model; <emph>Npar</emph>: number of parameters; Max <emph>z</emph>(<emph>r</emph>): maximum <emph>z</emph> score for r; Max. <emph>z</emph>(<emph>l</emph>): maximum <emph>z</emph> score for <emph>l</emph>. Adjusted <emph>p</emph>-values of Max. <emph>z</emph>(<emph>r</emph>) are indicated in the parentheses. Adjusted <emph>p</emph>-values are based on the Holm method.</p> <p>Since smaller BIC values indicate better model fit, Table 3 shows that the DINA (BIC = 1,966,743.99) and DINO (BIC = 1,966,743.99) models for quantitative skill and the R-RUM model (BIC = 1,976,088.06) provided the best fit among the models. However, it is important to note that the differences in goodness-of-fit indices between the models were very small, less than three. The R-RUM had the smallest maximum <emph>z</emph>-scores related to the <emph>r</emph> and <emph>l</emph> statistics among models. Overall, all these maximum <emph>z</emph>-scores were significant at the level of.001, showing that the test-takers' mastery profiles of the four skills of QVS could be measured by applying the CDM models.</p> <p>Furthermore, we investigated classification accuracy at both the test and skill levels to evaluate the fit and validity of applying CDM models to QVS. Test-level accuracy was.83 for quantitative skill and.79 for verbal skill, indicating that test-takers had an 83% chance of being correctly classified into the appropriate latent class for quantitative skill and a 79% chance for verbal skill. Table 4 presents classification accuracy at the skill level, showing high accuracy values ranging from.92 to.96 for quantitative skills and from.90 to.94 for verbal skills. Therefore, the results demonstrated that the application of CDM models to examine QVS in LSA tests was appropriate.</p> <p>Table 4. SweSAT Qualitative and Verbal Skills Mastery Proportion, Classification Accuracy, and Correlation.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left" rowspan="2">Skill</th><th align="left" rowspan="2">Definition</th><th align="left" rowspan="2">Mastery proportion</th><th align="left" rowspan="2">Classification accuracy</th><th align="left" colspan="3">Correlation</th></tr><tr><th align="left">A1</th><th align="left">A2</th><th align="left">A3</th></tr></thead><tbody><tr><td>Quantitative</td><td /><td /><td /><td /><td /><td /></tr><tr><td> A1</td><td>Mathematical problem-solving</td><td>.52</td><td>.96</td><td /><td /><td /></tr><tr><td> A2</td><td>Quantitative comparison</td><td>.50</td><td>.95</td><td>.99</td><td /><td /></tr><tr><td> A3</td><td>Data sufficiency</td><td>.53</td><td>.92</td><td>.83</td><td>.84</td><td /></tr><tr><td> A4</td><td>Diagrams, tables and maps</td><td>.40</td><td>.93</td><td>.67</td><td>.71</td><td>.86</td></tr><tr><td>Verbal</td><td /><td /><td /><td /><td /><td /></tr><tr><td> A1</td><td>Vocabulary</td><td>.45</td><td>.90</td><td /><td /><td /></tr><tr><td> A2</td><td>Swedish reading comprehension</td><td>.49</td><td>.91</td><td>.87</td><td /><td /></tr><tr><td> A3</td><td>Sentence completion</td><td>.48</td><td>.92</td><td>.95</td><td>.91</td><td /></tr><tr><td> A4</td><td>English reading comprehension</td><td>.50</td><td>.94</td><td>.69</td><td>.88</td><td>.84</td></tr></tbody></table> </ephtml> </p> <p>4 <emph>Note:</emph> All correlations are significant at the.01 level (2-tailed).</p> <hd id="AN0193364109-19">Examinee's Mastery Profiles</hd> <p>The expected a posteriori (EAP) was used to evaluate mastery profiles. Test-takers acquired a given skill if their mastery probability of that skill was higher than 0.50 ([<reflink idref="bib17" id="ref78">17</reflink>]; [<reflink idref="bib25" id="ref79">25</reflink>]; [<reflink idref="bib44" id="ref80">44</reflink>]). Table 4 indicates the mastery proportion and correlation between skills of QVS. The mastery proportion for quantitative skill was higher than for verbal skill, ranging from 0.40 to 0.52, while verbal skill ranged from 0.45 to 0.50. The quantitative skill A4 showed the lowest mastery proportion (0.40), indicating that test-takers were relatively weak at diagrams, tables and maps. In contrast, test takers demonstrated stronger mastery in DS (A3 = 0.53), mathematical problem-solving (A1 = 0.52), and QC (A2 = 0.50) in the quantitative part. Although the overall mastery values for the test-takers' verbal skill were higher than for their performance on diagrams, tables and maps in the quantitative skill, the highest value for verbal skill was in A4 (0.50), indicating that test-takers acquired ERC profoundly compared to other verbal skills. A1 showed the lowest mastery proportion (0.45), suggesting that test-takers were weak in vocabulary. In contrast, A2 (Swedish reading comprehension) and A3 (sentence completion) had higher mastery proportions (0.49 and 0.48, respectively). Thus, the test-takers' weak vocabulary knowledge did not significantly affect their performance in Swedish reading comprehension and sentence completion.</p> <p>Table 4 also shows phi correlations among QVS skills, indicating strong correlations between the skills in both quantitative and verbal parts, ranging from 0.67 to 0.99 in the quantitative part and from 0.69 to 0.95 in the verbal part. Notably, skills A1 and A2 in the quantitative part, and skills A1 and A3 in the verbal part, had the strongest correlations (0.99 and 0.95, respectively). However, skills A1 and A3 in both the quantitative and verbal parts had the lowest correlations, 0.67 and 0.69, respectively.</p> <p>Table 5 shows the test-takers' results across 16 mastery profiles and highlights both frequent and uncommon QVS among SweSAT participants. Nearly 37.36% of test-takers did not acquire any quantitative skill (0000), while the percentage of non-masters in verbal skills was higher at 39.52%. However, 32.38% of participants in the quantitative part and 35.28% in the verbal part mastered all QVS (1111). Test-takers were classified according to 16 mastery profiles showing that the most prevailed profiles were the 1110 profile (10.16% prevalence) in the quantitative part and the 0001 profile (4.70% prevalence) in the verbal part among SweSAT participants.</p> <p>Table 5. Mastery Profile of QVS and Prevalence.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /></colgroup><thead><tr><th align="left">Quantitative mastery profile</th><th align="left">Prevalence (%)</th><th align="left">Verbal mastery profile</th><th align="left">Prevalence (%)</th></tr></thead><tbody><tr><td>0000</td><td>37.36</td><td>0000</td><td>39.52</td></tr><tr><td>1000</td><td>0.97</td><td>1000</td><td>0.78</td></tr><tr><td>0100</td><td>0.28</td><td>0100</td><td>1.64</td></tr><tr><td>0010</td><td>4.76</td><td>0010</td><td>0.24</td></tr><tr><td>0001</td><td>1.79</td><td>0001</td><td>4.70</td></tr><tr><td>1100</td><td>5.02</td><td>1100</td><td>1.48</td></tr><tr><td>1010</td><td>1.10</td><td>1010</td><td>2.71</td></tr><tr><td>1001</td><td>0.52</td><td>1001</td><td>0.59</td></tr><tr><td>0110</td><td>0.66</td><td>0110</td><td>0.79</td></tr><tr><td>0101</td><td>0.11</td><td>0101</td><td>3.59</td></tr><tr><td>0011</td><td>3.13</td><td>0011</td><td>1.52</td></tr><tr><td>1110</td><td>10.16</td><td>1110</td><td>3.37</td></tr><tr><td>1101</td><td>1.18</td><td>1101</td><td>0.16</td></tr><tr><td>1011</td><td>0.44</td><td>1011</td><td>1.00</td></tr><tr><td>0111</td><td>0.14</td><td>0111</td><td>2.64</td></tr><tr><td>1111</td><td>32.38</td><td>1111</td><td>35.28</td></tr></tbody></table> </ephtml> </p> <p>5 <emph>Note:</emph> All correlations are significant at the.01 level (2-tailed).</p> <hd id="AN0193364109-20">Measurement Invariance</hd> <p>DIF was examined to investigate measurement invariance related to sex, age and educational level. Appendix B shows that all 80 items of QVS did not exhibit DIF associated with sex, age and educational level. Thus, the QVS items were free from bias across sex, age and educational level.</p> <hd id="AN0193364109-21">Background Information and QVS Achievement</hd> <p></p> <hd id="AN0193364109-22">Simple Logistic Regression Results</hd> <p>To examine the relationship between test-takers' background information and their achievement in each quantitative and verbal skill, latent logistic regression was performed. Results are presented in Tables 6 and 7. First, a simple logistic regression analysis was conducted to investigate the relationship between each covariate and each QVS. The results indicate that the odds ratio for sex was greater than 1. Specifically, the odds ratio for females was higher (1.31–1.62 = <emph>e</emph><sups>0.27×0.48</sups>, <emph>p</emph> <.001) for both QVS, indicating a relatively strong association between sex and mastery of QVS. A relatively significant association was observed between the reference age group (up to 20 years old) and all skills in both the quantitative and verbal subskills (1.47–1.85 = <emph>e</emph><sups>0.38×0.62</sups> and 1.47–1.64 = <emph>e</emph><sups>0.39×0.49</sups>, respectively). However, the patterns differed between the quantitative and verbal parts. For example, if the odd ratio for the up to 20 years old group was the highest (1.85 = <emph>e</emph><sups>0.62</sups>) in A1 skill and the lowest in A4 (1.47 = <emph>e</emph><sups>0.38</sups>) in quantitative part, the odd ratio for the same age group was the lowest (1.47 = <emph>e</emph><sups>0.39</sups>) in A1 skill and the highest (1.64 = <emph>e</emph><sups>0.49</sups>) in A4 in verbal part. The log of the odds of the test-takers' age significantly impacted their mastery status of four QVS skills, with effect sizes ranging from −0.16 to −0.02 in the quantitative part (Table 6) and from −0.00 to 0.26 in the verbal part (Table 7). These results revealed a negative association with age (21–40 years old or above) in the quantitative part, indicating that younger test-takers performed better in A1, A2 and A3 skills. Conversely, there was a positive association with age (21–40 years old or above) in these three verbal subskills, indicating that the older test-takers performed better in verbal skills. For example, if the 40 years old or above age group had difficulties in mathematical problem-solving (A1) skills (0.80 = <emph>e</emph><sups>−0.23</sups>, <emph>d</emph> = −0.12) in quantitative part, they had relatively higher results in vocabulary (A1) skill (1.59 = <emph>e</emph><sups>0.46</sups>, <emph>d</emph> = 0.26) in verbal part. However, there was no difference between odd ratios of age groups related to A4 skill in both quantitative and verbal parts.</p> <p>Table 6. Simple Logistic Regression, Odds and Odds Ratios, and Effect Sizes of Quantitative Skills.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /></colgroup><thead><tr><th align="left" rowspan="2" /><th align="left" colspan="3">A1</th><th align="left" colspan="3">A2</th><th align="left" colspan="3">A3</th><th align="left" colspan="3">A4</th></tr><tr><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th></tr></thead><tbody><tr><td>Gender</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Male</td><td>0.15</td><td>1.26</td><td>.13</td><td>0.16<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.17</td><td>.09</td><td>0.17<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.18</td><td>.09</td><td>0.19<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.21</td><td>.11</td></tr><tr><td>Age</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> 21–24 years old</td><td>−0.16<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.85</td><td>−.09</td><td>−0.16<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.85</td><td>−.09</td><td>−0.11<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.90</td><td>−.06</td><td>0.08<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.92</td><td>−.05</td></tr><tr><td> 25–29 years old</td><td>−0.25<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.78</td><td>−.13</td><td>−0.24<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.79</td><td>−.13</td><td>−0.19<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.82</td><td>−.11</td><td>−0.13<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.88</td><td>−.07</td></tr><tr><td> 30–39 years old</td><td>−0.29<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.75</td><td>−.16</td><td>−0.28<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.76</td><td>−.15</td><td>−0.17<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.84</td><td>−.10</td><td>−0.09<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.91</td><td>−.05</td></tr><tr><td> 40 years old or above</td><td>−0.23<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.80</td><td>−.12</td><td>−0.23<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.80</td><td>−.12</td><td>−0.13<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.87</td><td>−.07</td><td>−0.05<xref ref-type="table-fn" rid="tfn7">**</xref></td><td>0.96</td><td>−.02</td></tr><tr><td>Educational level</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Compulsory school</td><td>−0.15<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.86</td><td>−.08</td><td>−0.15<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.86</td><td>−.08</td><td>0.00</td><td>1.00</td><td>.00</td><td>−0.14<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.87</td><td>−.07</td></tr><tr><td> Folk high school</td><td>−0.26<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.77</td><td>−.14</td><td>−0.25<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.78</td><td>−.13</td><td>−0.06</td><td>0.94</td><td>−.03</td><td>−0.18<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.84</td><td>−.10</td></tr><tr><td> Upper secondary school, less or 2 years</td><td>−0.34<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.72</td><td>−.18</td><td>−0.32<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.72</td><td>−.19</td><td>−0.13<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.88</td><td>−.07</td><td>−0.21<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>0.81</td><td>−.12</td></tr><tr><td> Upper secondary school, 3 or more years</td><td>0.05<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.05</td><td>.03</td><td>0.04<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.04</td><td>.02</td><td>0.19<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.21</td><td>.11</td><td>−0.00</td><td>1.00</td><td>.00</td></tr><tr><td> Postsecondary education, less than 3 years</td><td>0.08<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.08</td><td>.04</td><td>0.08<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.06</td><td>.03</td><td>0.27<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.31</td><td>.15</td><td>0.09<xref ref-type="table-fn" rid="tfn8">***</xref></td><td>1.10</td><td>.05</td></tr></tbody></table> </ephtml> </p> <ulist> <item>6 <emph>Note:</emph> Intercepts of regression models are omitted to save space; <emph>B</emph>: unstandardized coefficients; ES: effect size.</item> <item>7 <emph>p</emph> <.01.</item> <item>8 <emph>p</emph> <.001.</item> </ulist> <p>Table 7. Simple Logistic Regression, Odds and Odds Ratios, and Effect Sizes of Verbal Skills.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /></colgroup><thead><tr><th align="left" rowspan="2" /><th align="left" colspan="3">A1</th><th align="left" colspan="3">A2</th><th align="left" colspan="3">A3</th><th align="left" colspan="3">A4</th></tr><tr><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th></tr></thead><tbody><tr><td><italic>Gender</italic></td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Male</td><td>−0.00</td><td>1.00</td><td>.00</td><td>0.07<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.07</td><td>.04</td><td>0.06<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.06</td><td>.03</td><td>0.16<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.17</td><td>.09</td></tr><tr><td><italic>Age</italic></td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> 21–24 years old</td><td>0.04<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.04</td><td>.02</td><td>0.01</td><td>1.01</td><td>.01</td><td>0.02<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.02</td><td>.01</td><td>−0.01</td><td>1.00</td><td>.00</td></tr><tr><td> 25–29 years old</td><td>0.14<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.15</td><td>.08</td><td>0.05<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.06</td><td>.03</td><td>0.12<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.12</td><td>.06</td><td>0.03<xref ref-type="table-fn" rid="tfn10">**</xref></td><td>1.04</td><td>.02</td></tr><tr><td> 30–39 years old</td><td>0.32<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.38</td><td>.18</td><td>0.06<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.18</td><td>.09</td><td>0.28<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.32</td><td>.15</td><td>0.09<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.09</td><td>.05</td></tr><tr><td> 40 years old or above</td><td>0.46<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.59</td><td>.26</td><td>0.26<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.29</td><td>.14</td><td>0.41<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.50</td><td>.22</td><td>0.16<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.18</td><td>.09</td></tr><tr><td><italic>Educational level</italic></td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Compulsory school</td><td>−0.24<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.79</td><td>−.13</td><td>−0.21<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.80</td><td>−.12</td><td>−0.22<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.80</td><td>−.12</td><td>−0.19<xref ref-type="table-fn" rid="tfn10">**</xref></td><td>0.82</td><td>−.11</td></tr><tr><td> Folk high school</td><td>−0.14<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.87</td><td>−.08</td><td>−0.14<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.87</td><td>−.08</td><td>−0.14<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.87</td><td>−.08</td><td>−0.12<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.89</td><td>−.06</td></tr><tr><td> Upper secondary school, less or 2 years</td><td>−0.21<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.81</td><td>−.12</td><td>−0.23<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.80</td><td>−.12</td><td>−0.21<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.81</td><td>−.12</td><td>−0.25<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.78</td><td>−.07</td></tr><tr><td> Upper secondary school, 3 or more years</td><td>−0.15<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.86</td><td>−.08</td><td>−0.10<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.90</td><td>−.06</td><td>−0.14<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.87</td><td>−.08</td><td>−0.09<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>0.91</td><td>−.05</td></tr><tr><td> Postsecondary education, less than 3 years</td><td>0.09<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.10</td><td>.05</td><td>0.13<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.14</td><td>.07</td><td>0.12<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.13</td><td>.07</td><td>0.11<xref ref-type="table-fn" rid="tfn11">***</xref></td><td>1.11</td><td>.06</td></tr></tbody></table> </ephtml> </p> <ulist> <item>9 <emph>Note:</emph> Intercepts of regression models are omitted to save space; <emph>B</emph>: unstandardized coefficients; ES: effect size.</item> <item>10 <emph>p</emph> <.01.</item> <item>11 <emph>p</emph> <.001.</item> </ulist> <p>Compared to postsecondary education of 3 or more years, compulsory, folk high schools, and upper secondary schools were negatively associated with all subskills in the quantitative part (Table 6). In contrast, only postsecondary education of less than 3 years showed a positive association with all subskills in the verbal part (Table 7). The log odds and effect sizes for being in upper secondary school of 3 or more years and for postsecondary education of less than 3 years, compared to postsecondary education of 3 or more years, were relatively largest for the A3 skill in the quantitative part (1.21 = <emph>e</emph><sups>0.19</sups>, <emph>d</emph> = 0.11 and 1.31 = <emph>e</emph><sups>0.27</sups>, <emph>d</emph> = 0.15, respectively). Although the odds ratio for the reference educational level (postsecondary education of 3 or more years) showed the most significant association with QVS among educational levels, this association was lower in all skills of the quantitative part, ranging from 1.14 to 1.68 (<emph>e</emph><sups>0.36</sups> to <emph>e</emph><sups>0.52</sups>), compared to the verbal part, which ranged from 1.74 to 1.78 (<emph>e</emph><sups>0.55</sups> to <emph>e</emph><sups>0.58</sups>). These findings indicate that the higher the educational level, the more test-takers would have a higher mastery profile and would perform better in the SweSAT test in both QVS, specifically in the quantitative part.</p> <hd id="AN0193364109-23">Multiple Logistic Regression Results</hd> <p>Multiple regression analysis was performed to understand how the covariates simultaneously impacted each quantitative and verbal skill. Since all background variables showed relatively significant associations with QVS in simple logistic regression results, we investigated the relationship between all background characteristics (gender, age and educational level) and QVS (Tables 8 and 9). The selection of covariates for inclusion in multiple logistic regression analysis was based on their associations with QVS from simple logistic regression analysis. Similarly, the findings from the simple logistic regression, gender had a significant relationship with each skill in both parts of QVS. However, there were small differences between simple and multiple regression analyses. While the log odds of age and educational level were negatively associated with quantitative skills (except for upper secondary school of 3 or more years and postsecondary education of less than 3 years, compared to the reference educational level of postsecondary education of 3 or more years), there was a positive association between age groups and verbal skills on the SweSAT, compared to a reference age group of 20 years old or younger, with odd ratios ranging from 1.00 to 1.56 (<emph>e</emph><sups>0.00</sups> to <emph>e</emph><sups>0.45</sups>). Older test-takers tended to have higher mastery levels in vocabulary skill.</p> <p>Table 8. Multiple Logistic Regression, Odds and Odds Ratios, and Effect Sizes of Quantitative Skills.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /></colgroup><thead><tr><th align="left" rowspan="2" /><th align="left" colspan="3">A1</th><th align="left" colspan="3">A2</th><th align="left" colspan="3">A3</th><th align="left" colspan="3">A4</th></tr><tr><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th></tr></thead><tbody><tr><td>Gender</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Male</td><td>0.15<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.16</td><td>.08</td><td>0.16<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.17</td><td>.09</td><td>0.17<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.18</td><td>.09</td><td>0.19<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.21</td><td>.11</td></tr><tr><td>Age</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> 21–24 years old</td><td>−0.17<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.85</td><td>−.09</td><td>−0.17<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.84</td><td>−.10</td><td>−0.12<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.89</td><td>−.06</td><td>−0.10<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.91</td><td>−.05</td></tr><tr><td> 25–29 years old</td><td>−0.25<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.78</td><td>−.14</td><td>−0.24<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.78</td><td>−.14</td><td>−0.20<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.82</td><td>−.11</td><td>−0.14<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.87</td><td>−.08</td></tr><tr><td> 30–39 years old</td><td>−0.31<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.74</td><td>−.17</td><td>−0.30<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.74</td><td>−.17</td><td>−0.19<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.83</td><td>−.10</td><td>−0.11<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.90</td><td>−.06</td></tr><tr><td> 40 years old or above</td><td>−0.26<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.77</td><td>−.14</td><td>−0.27<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.76</td><td>−.15</td><td>−0.16<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.85</td><td>−.09</td><td>−0.07<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.93</td><td>−.04</td></tr><tr><td>Educational level</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Compulsory school</td><td>−0.27<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.77</td><td>−.14</td><td>−0.26<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.77</td><td>−.14</td><td>−0.25<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.78</td><td>−.14</td><td>−0.20<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.82</td><td>−.11</td></tr><tr><td> Folk high school</td><td>−0.26<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.77</td><td>−.14</td><td>−0.24<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.79</td><td>−.13</td><td>−0.21<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.81</td><td>−.12</td><td>−0.15<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.86</td><td>−.08</td></tr><tr><td> Upper secondary school, less or 2 years</td><td>−0.38<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.69</td><td>−.21</td><td>−0.36<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.70</td><td>−.20</td><td>−0.32<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.73</td><td>−.17</td><td>−0.22<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.81</td><td>−.12</td></tr><tr><td> Upper secondary school, 3 or more years</td><td>0.10<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.91</td><td>−.05</td><td>0.10<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.91</td><td>−.05</td><td>−0.06<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.94</td><td>−.03</td><td>0.05<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>0.95</td><td>−.03</td></tr><tr><td> Postsecondary education, less than 3 years</td><td>0.04<xref ref-type="table-fn" rid="tfn13">**</xref></td><td>1.04</td><td>.02</td><td>0.04<xref ref-type="table-fn" rid="tfn13">**</xref></td><td>1.04</td><td>.02</td><td>0.08<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.07</td><td>.04</td><td>0.09<xref ref-type="table-fn" rid="tfn14">***</xref></td><td>1.09</td><td>.05</td></tr></tbody></table> </ephtml> </p> <ulist> <item>12 <emph>Note:</emph> Intercepts of regression models are omitted to save space; <emph>B</emph>: unstandardized coefficient; ES: effect size.</item> <item>13 <emph>p</emph> <.01.</item> <item>14 <emph>p</emph> <.001.</item> </ulist> <p>Table 9. Multiple Logistic Regression, Odds and Odds Ratios, and Effect Sizes of Verbal Skills.</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /><col align="char" char="." /></colgroup><thead><tr><th align="left" rowspan="2" /><th align="left" colspan="3">A1</th><th align="left" colspan="3">A2</th><th align="left" colspan="3">A3</th><th align="left" colspan="3">A4</th></tr><tr><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th><th align="left"><italic>B</italic></th><th align="left">Odds ratio</th><th align="left">ES</th></tr></thead><tbody><tr><td>Gender</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Male</td><td>−0.00</td><td>1.00</td><td>.00</td><td>0.07<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.07</td><td>.04</td><td>0.06<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.07</td><td>.04</td><td>0.16<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.17</td><td>.09</td></tr><tr><td>Age</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> 21–24 years old</td><td>0.02<xref ref-type="table-fn" rid="tfn17">**</xref></td><td>1.02</td><td>.01</td><td>−0.02<xref ref-type="table-fn" rid="tfn16">*</xref></td><td>0.98</td><td>−.01</td><td>0.00</td><td>1.00</td><td>.00</td><td>−0.03<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.97</td><td>−.02</td></tr><tr><td> 25–29 years old</td><td>0.13<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.14</td><td>.07</td><td>0.04<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.04</td><td>.02</td><td>0.10<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.11</td><td>.06</td><td>0.02</td><td>1.01</td><td>.01</td></tr><tr><td> 30–39 years old</td><td>0.30<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.36</td><td>.17</td><td>0.14<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.15</td><td>.08</td><td>0.26<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.29</td><td>.14</td><td>0.06<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.06</td><td>.03</td></tr><tr><td> 40 years old or above</td><td>0.45<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.56</td><td>.25</td><td>0.22<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.25</td><td>.12</td><td>0.38<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.47</td><td>.21</td><td>0.13<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.14</td><td>.07</td></tr><tr><td>Educational level</td><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td> Compulsory school</td><td>−0.12<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.13</td><td>.07</td><td>−0.17<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.84</td><td>−.10</td><td>−0.13<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.88</td><td>−.07</td><td>−0.17<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.84</td><td>−.10</td></tr><tr><td> Folk high school</td><td>−0.07<xref ref-type="table-fn" rid="tfn16">*</xref></td><td>1.20</td><td>.10</td><td>−0.10<xref ref-type="table-fn" rid="tfn17">**</xref></td><td>0.91</td><td>−.05</td><td>−0.08<xref ref-type="table-fn" rid="tfn16">*</xref></td><td>0.93</td><td>−.04</td><td>−0.08<xref ref-type="table-fn" rid="tfn17">**</xref></td><td>0.92</td><td>−.05</td></tr><tr><td> Upper secondary school, less or 2 years</td><td>−0.12<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.14</td><td>.07</td><td>−0.18<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.84</td><td>−.10</td><td>−0.13<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.88</td><td>−.07</td><td>−0.21<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.81</td><td>−.12</td></tr><tr><td> Upper secondary school, 3 or more years</td><td>−0.01</td><td>1.27</td><td>.13</td><td>−0.04<xref ref-type="table-fn" rid="tfn17">**</xref></td><td>0.97</td><td>.02</td><td>−0.01</td><td>0.99</td><td>−.01</td><td>−0.05<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>0.95</td><td>−.03</td></tr><tr><td> Postsecondary education, less than 3 years</td><td>0.18<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.54</td><td>.24</td><td>0.18<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.19</td><td>.10</td><td>0.20<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.22</td><td>.11</td><td>0.14<xref ref-type="table-fn" rid="tfn18">***</xref></td><td>1.15</td><td>.08</td></tr></tbody></table> </ephtml> </p> <ulist> <item>15 <emph>Note:</emph> Intercepts of regression models are omitted to save space; <emph>B</emph>: unstandardized coefficient; ES: effect size.</item> <item>16 <emph>p</emph> <.05.</item> <item>17 <emph>p</emph> <.01.</item> <item>18 <emph>p</emph> <.001.</item> </ulist> <p>Further, the results of multiple logistic regression on educational level related to all skills were similar to those from the simple logistic regression, showing that postsecondary educational levels had a relatively stronger association with all skills in the verbal part. To summarize regardless of whether the skill is quantitative or verbal, a test-taker with a higher educational level is more likely to be successful on the SweSAT.</p> <hd id="AN0193364109-24">Discussion</hd> <p>The quest for effective analysis methods to provide detailed insights into test-takers' performance and ensure the validity of high-stakes tests has been ongoing for decades. Detailed insights into test-takers are essential for developing programs that offer equitable support to high-ability students, regardless of their background information, allowing them to maximize their potential through opportunities that extend beyond traditional boundaries. This necessity emphasizes the call for substantial systemic reforms to correct existing inequities. Thus, it is imperative to choose the appropriate analytical method to explore the relationship between QVS and background characteristics. The application of a suitable method can provide valuable information about test-takers to their instructors, which can enhance equal access to advanced educational opportunities and address the underlying factors that contribute to gaps in opportunity, excellence, and equity.</p> <p>This study contributes to this body of research by applying a general CDM framework to analyze the performance of QVS in the SweSAT. Utilizing a three-step approach, we examined the relationships between test-takers' background characteristics and their performance across various subskills within both quantitative and verbal domains.</p> <p>The results, which showed acceptable classification accuracy at the test level and high classification accuracies at the skill level, provided evidence for the adequate fit of CDMs to QVS data. This indicates that the application of CDMs to the assessment of QVS is both viable and appropriate. CDMs enable test developers to obtain diagnostic information about each test-taker's mastery profiles and the mastering probabilities of each skill estimated from person parameters. This information can contribute to the development of SweSAT measurements. This study found high mastery proportions—above 50%—for three quantitative skills on the SweSAT, compared to the verbal part, where mastery proportion were below 50%. This suggests that the verbal part was more challenging for SweSAT test-takers. The lowest mastery proportion was found for the quantitative skill A4 (diagrams, tables and maps). Mastery proportions for English and Swedish reading comprehension were the second and third lowest among the SweSAT skills. The lowest mastery proportion among the verbal skills was observed for A1 (vocabulary). A possible reason why test-takers had a higher mastery proportion in both reading comprehensions than vocabulary could be guessing vocabulary and inferring meanings from a text ([<reflink idref="bib12" id="ref81">12</reflink>]).</p> <p>Further, this study conducted a DIF analysis across various groups of test-takers related to sex, age and educational level of SweSAT in a large-scale study context and revealed no items exhibiting sex, age, educational level-related bias, suggesting that observed score differences reflect true performance variation rather than measurement bias. All 80 studied items were free of DIF, providing evidence for the validity and fairness of the high-stakes SweSAT test and ensuring the generalizability of the instrument across sex, age and various educational levels.</p> <p>The findings of simple latent logistic regression indicated various associations between test-takers' background characteristics and mastery of QVS. Although, this study found a relationship between gender and QVS, which aligns with [<reflink idref="bib36" id="ref82">36</reflink>] and [<reflink idref="bib13" id="ref83">13</reflink>] findings, the results of simple logistic analysis contrast with [<reflink idref="bib13" id="ref84">13</reflink>] study, indicating a high odds ratio for women. This finding reveals the necessity of further investigations into covariate interactions. [<reflink idref="bib11" id="ref85">11</reflink>] found that age moderated the effect of gender on deep learning.</p> <p>Further, the test-takers' age was significantly associated with their mastery of QVS in this study. This study found that the higher the age, the greater the mastery of vocabulary in the verbal part, whereas increasing age had less influence on mastery of DS and DTM, which is consistent with the findings of [<reflink idref="bib36" id="ref86">36</reflink>]. Further, the finding that vocabulary mastery increased monotonically with age, whereas quantitative-reasoning mastery decreased, mirrors longitudinal evidence of differential cognitive ageing ([<reflink idref="bib34" id="ref87">34</reflink>]). Because no items exhibited age-related DIF, these trends likely reflect differential practice rather than measurement bias introduced by the test. Regardless of age group, test-takers struggled with the quantitative part of the SweSAT, while higher educational levels were associated with a higher mastery profile in the verbal part. These findings highlight the need for further investigation into covariate interactions.</p> <p>Next, the findings of this study showed that the educational level was positively associated with the performance. This finding is also reported by Stage and Ögen's study ([<reflink idref="bib36" id="ref88">36</reflink>]). Thus, test-takers with a higher educational level are more likely to succeed on high-stakes tests such as SweSAT.</p> <p>The present study also conducted multiple logistic regression analyses to explore how covariates simultaneously related to QVS. Similar to the result of the simple logistic regression, the findings from the multiple logistic regression showed varying relationships between background characteristics and test-takers' mastery profiles.</p> <p>The difference between previous studies that used general test scores, and this study is the provision of fine-grained information about test-takers' weaknesses and strengths in high-stakes LSA tests, such as the SweSAT, for test developers and educators. CDMs with latent logistic regression can give insights into the associations between the test-takers' mastery profiles and background characteristics, which differ among skills. Following [<reflink idref="bib27" id="ref89">27</reflink>] study, this study explored relationships between mastery profiles and covariates by applying Iaconangelo and de la Torre's ([<reflink idref="bib18" id="ref90">18</reflink>]) method at the test-level. The skill-level associations between background variables and mastery levels can be an object of future study. This study contributed to high-stakes educational measurement research by providing evidence of the application of CDMs with covariates in a large-scale testing context. However, this study has limitations. The first limitation is that it explored relationships between mastery profiles and covariates. Future studies could examine interaction effects to better understand how the interaction of gender and age is associated with test-takers' mastery profiles. The second limitation is that in this study, the researchers did not examine the affective variables of test-takers; factors such as interest, motivation and engagement, which could influence their participation in the SweSAT and might have a larger impact on their mastery profiles compared to background characteristics. It would also be insightful to study the interaction effect between non-cognitive skills and background variables in relation to mastery profiles. Moreover, some studies revealed that the range of response time where between-condition differences in the effect of response time on the probability of a correct response were accurate ([<reflink idref="bib5" id="ref91">5</reflink>]). The integration of response time to the CDM models applying three-step latent class analysis can provide more insights about the test-taker.</p> <hd id="AN0193364109-25">Appendix A The Q -Matrix</hd> <p></p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Quantitative items</th><th align="left">A1</th><th align="left">A2</th><th align="left">A3</th><th align="left">A4</th><th align="left">Verbal items</th><th align="left">A1</th><th align="left">A2</th><th align="left">A3</th><th align="left">A4</th></tr></thead><tbody><tr><td><p>1</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>2</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>2</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>3</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>3</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>4</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>4</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>5</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>5</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>6</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>6</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>7</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>7</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>8</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>8</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>9</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>9</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>10</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>10</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>11</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>11</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>12</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>12</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>13</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>13</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>14</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>14</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>15</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>15</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>16</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>16</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>17</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>17</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>18</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>18</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>19</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>19</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>20</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>20</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td></tr><tr><td><p>21</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>21</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>22</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>22</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>23</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>0</p></td><td><p>23</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>24</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>24</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>25</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>25</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>26</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>26</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>27</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>27</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>28</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>28</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>29</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td><td><p>29</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>30</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>30</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>0</p></td></tr><tr><td><p>31</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>31</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>32</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>32</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>33</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>33</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>34</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>34</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>35</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>35</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>36</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>36</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>37</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>37</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>38</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>38</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>39</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>39</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr><tr><td><p>40</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td><td><p>40</p></td><td><p>0</p></td><td><p>0</p></td><td><p>0</p></td><td><p>1</p></td></tr></tbody></table> </ephtml> </p> <p>19 <emph>Note</emph>. Quantitative A1 = Mathematical problem-solving; A2 = Quantitative comparison; A3 = Data sufficiency; A4 = Diagrams, tables and maps; Verbal A1 = Vocabulary; A2 = Swedish reading comprehension; A3 = Sentence completion; A4 = English reading comprehension.</p> <hd id="AN0193364109-26">Appendix B Differential Item Functioning</hd> <p></p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /><col align="left" /></colgroup><thead><tr><th align="left">Item</th><th align="left" colspan="3">Sex</th><th align="left" colspan="3">Age</th><th align="left" colspan="3">Education level</th></tr><tr><th align="left" /><th align="left">Wald stat.</th><th align="left"><italic>p</italic></th><th align="left"><italic>df</italic></th><th align="left">Wald stat.</th><th align="left"><italic>p</italic></th><th align="left"><italic>df</italic></th><th align="left">Wald stat.</th><th align="left"><italic>p</italic></th><th align="left"><italic>df</italic></th></tr></thead><tbody><tr><td><p>Q</p></td><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td><p>1</p></td><td><p>16.31</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>73.26</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>46.52</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>2</p></td><td><p>12.99</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>13.80</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.43</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>3</p></td><td><p>49.25</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>116.10</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>63.47</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>4</p></td><td><p>44.67</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>121.87</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>69.63</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>5</p></td><td><p>135.98</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>46.29</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>52.29</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>6</p></td><td><p>64.90</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>46.01</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.33</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>7</p></td><td><p>72.05</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>335.36</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>103.41</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>8</p></td><td><p>38.41</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>260.20</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>76.84</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>9</p></td><td><p>98.60</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>12.06</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>20.11</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>10</p></td><td><p>102.55</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>240.04</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>149.24</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>11</p></td><td><p>46.11</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>11.52</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>15.07</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>12</p></td><td><p>69.29</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>112.19</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>39.94</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>13</p></td><td><p>164.07</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>390.06</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>129.98</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>14</p></td><td><p>154.41</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>130.28</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>57.66</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>15</p></td><td><p>219.27</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>419.20</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>153.78</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>16</p></td><td><p>48.40</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>23.10</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>22.98</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>17</p></td><td><p>65.81</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>50.67</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>27.12</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>18</p></td><td><p>99.25</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>266.16</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>141.49</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>19</p></td><td><p>821.51</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>101.78</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>57.36</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>20</p></td><td><p>49.38</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>79.48</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>56.36</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>21</p></td><td><p>33.11</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>27.43</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>41.85</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>22</p></td><td><p>25.96</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>29.28</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>31.24</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>23</p></td><td><p>42.15</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>59.51</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>34.99</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>24</p></td><td><p>0.59</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>117.12</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>63.01</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>25</p></td><td><p>40.91</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>42.31</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.94</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>26</p></td><td><p>103.89</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>31.69</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>27.71</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>27</p></td><td><p>4.10</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>55.24</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>23.30</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>28</p></td><td><p>4.28</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>96.23</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>27.68</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>29</p></td><td><p>105.39</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>140.05</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>92.95</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>30</p></td><td><p>21.00</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>64.13</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>25.71</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>31</p></td><td><p>51.12</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>194.94</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>56.01</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>32</p></td><td><p>99.77</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>111.74</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>51.19</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>33</p></td><td><p>149.68</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>58.79</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.67</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>34</p></td><td><p>19.37</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>151.08</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>42.17</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>35</p></td><td><p>2.76</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>88.50</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>55.01</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>36</p></td><td><p>169.07</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>110.99</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>56.04</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>37</p></td><td><p>82.69</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>110.18</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>53.75</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>38</p></td><td><p>68.03</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>41.89</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>45.08</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>39</p></td><td><p>52.17</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>20.57</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.31</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>40</p></td><td><p>78.64</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>40.83</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>36.34</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>V</p></td><td /><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td><p>1</p></td><td><p>303.10</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>171.93</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>90.08</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>2</p></td><td><p>829.60</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>374.49</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>103.85</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>3</p></td><td><p>7.18</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>281.40</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>116.68</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>4</p></td><td><p>58.29</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>164.18</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>71.17</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>5</p></td><td><p>113.11</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>245.60</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>118.43</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>6</p></td><td><p>98.09</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>157.42</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>99.70</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>7</p></td><td><p>296.62</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>154.81</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>86.03</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>8</p></td><td><p>3.35</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>213.82</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>184.46</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>9</p></td><td><p>27.37</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>147.15</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>94.25</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>10</p></td><td><p>87.89</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>817.12</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>241.43</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>11</p></td><td><p>3.30</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>6.58</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>34.09</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>12</p></td><td><p>3.10</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>29.10</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>19.44</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>13</p></td><td><p>94.90</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>61.44</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>21.33</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>14</p></td><td><p>356.17</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>113.45</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>60.81</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>15</p></td><td><p>271.331</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>117.92</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>26.99</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>16</p></td><td><p>10.37</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>489.93</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>150.87</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>17</p></td><td><p>33.67</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>74.81</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>34.45</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>18</p></td><td><p>19.36</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>51.34</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>43.52</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>19</p></td><td><p>6.94</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>38.84</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>31.97</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>20</p></td><td><p>4.61</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>39.53</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>45.43</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>21</p></td><td><p>331.61</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>45.41</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>49.10</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>22</p></td><td><p>48.32</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>40.85</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>52.21</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>23</p></td><td><p>12.69</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>272.67</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>83.77</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>24</p></td><td><p>100.04</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>23.33</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>34.18</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>25</p></td><td><p>46.40</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>439.27</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>141.21</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>26</p></td><td><p>19.90</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>290.63</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>135.35</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>27</p></td><td><p>198.25</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>179.72</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>55.08</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>28</p></td><td><p>37.04</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>635.49</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>334.64</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>29</p></td><td><p>1043.46</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>702.89</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>361.90</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>30</p></td><td><p>442.73</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>475.29</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>605.24</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>31</p></td><td><p>86.29</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>200.89</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>37.40</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>32</p></td><td><p>1081.12</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>42.02</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>23.29</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>33</p></td><td><p>71.22</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>358.63</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>118.50</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>34</p></td><td><p>140.76</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>50.90</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>22.20</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>35</p></td><td><p>16.94</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>298.42</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>114.82</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>36</p></td><td><p>446.38</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>267.60</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>79.61</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>37</p></td><td><p>171.33</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>220.14</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>96.38</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>38</p></td><td><p>248.14</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>278.99</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>58.13</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>39</p></td><td><p>125.20</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>94.59</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>22.07</p></td><td><p>1.00</p></td><td><p>12</p></td></tr><tr><td><p>40</p></td><td><p>408.10</p></td><td><p>1.00</p></td><td><p>2</p></td><td><p>274.78</p></td><td><p>1.00</p></td><td><p>8</p></td><td><p>106.31</p></td><td><p>1.00</p></td><td><p>12</p></td></tr></tbody></table> </ephtml> </p> <p>20 <emph>Note</emph>. Wald stat. = Wald statistic; <emph>p</emph>-values were adjusted based on the Holm method; an item is flagged as a DIF item if <emph>p</emph> <.05.</p> <ref id="AN0193364109-27"> <title> References </title> <blist> <bibl id="bib1" idref="ref69" type="bt">1</bibl> <bibtext> Akaike H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), 716–723. https://doi.org/10.1109/TAC.1974.1100705</bibtext> </blist> <blist> <bibl id="bib2" idref="ref1" type="bt">2</bibl> <bibtext> Bastedo M. N., Glasener K. M., Deane K. C., Bowman N. A. (2022). Contextualizing the SAT: Experimental evidence on college admission recommendations for low-SES applicants. Educational Policy, 36(2), 282–311. https://doi.org/10.1177/0895904819874752</bibtext> </blist> <blist> <bibl id="bib3" idref="ref58" type="bt">3</bibl> <bibtext> Bolt D., Chen H., DiBello L., Hartz S., Henson R., Roussos L., Stout W., Templin J. (2008). The Arpeggio Suite: Software for cognitive skills diagnostic assessment [Computer software and manual]. Assessment Systems.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref44" type="bt">4</bibl> <bibtext> Chen J., de la Torre J., Zhang Z. (2013). Relative and absolute fit evaluation in cognitive diagnosis modeling. Journal of Educational Measurement, 50(2), 123–140. https://doi.org/10.1111/j.1745-3984.2012.00185.x</bibtext> </blist> <blist> <bibl id="bib5" idref="ref91" type="bt">5</bibl> <bibtext> Cho S.-J., Goodwin A. P., Naveiras M., Salas J. (2024). Differential and functional response time item analysis: An application to understanding paper versus digital reading processes. Journal of Educational Measurement, 61(2), 219–251. https://doi.org/10.1111/jedm.12389</bibtext> </blist> <blist> <bibl id="bib6" idref="ref76" type="bt">6</bibl> <bibtext> Cohen J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Routledge.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref54" type="bt">7</bibl> <bibtext> de la Torre J. (2009). DINA model and parameter estimation: A didactic. Journal of Educational and Behavioral Statistics, 34(1), 115–130. https://doi.org/10.3102/1076998607309474</bibtext> </blist> <blist> <bibl id="bib8" idref="ref45" type="bt">8</bibl> <bibtext> de la Torre J. (2011). The generalized DINA model framework. Psychometrika, 76(2), 179–199. https://doi.org/10.1007/s11336-011-9207-7</bibtext> </blist> <blist> <bibl id="bib9" idref="ref51" type="bt">9</bibl> <bibtext> de la Torre J., Minchen N. (2014). Cognitively diagnostic assessments and the cognitive diagnosis model framework. Psicologia Educativa, 20(2), 89–97. https://doi.org/10.1016/j.pse.2014.11.001</bibtext> </blist> <blist> <bibtext> de la Torre J., van der Ark L. A., Rossi G. (2018). Analysis of clinical data from a cognitive diagnosis modeling framework. Measurement and Evaluation in Counseling and Development, 51(4), 281–296. https://doi.org/10.1080/07481756.2017.1327286</bibtext> </blist> <blist> <bibtext> Douglas H. E., Rubin M., Scevak J., Southgate E., Macqueen S., Richardson J. T. E. (2020). Older women, deeper learning: age and gender interact to predict learning approach and academic achievement at university. Frontiers in Education, 5, Article 158. https://doi.org/10.3389/feduc.2020.00158</bibtext> </blist> <blist> <bibtext> Gebremeskel T. A., Bachore M. M., Bushisho E. W. (2024). The effects of multiple intelligence-based reading tasks on EFL students reading skills achievements: The case of university students in Ethiopia. Heliyon, 10(13), e33591. https://doi.org/10.1016/j.heliyon.2024.e33591</bibtext> </blist> <blist> <bibtext> Graetz G., Karimi A. (2022). Gender gap variation across assessment types: Explanations and implications. Economics of Education Review, 91, 102313. https://doi.org/10.1016/j.econedurev.2022.102313</bibtext> </blist> <blist> <bibtext> Hannon B., McNaughton-Cassill M. (2011). SAT performance: Understanding the contributions of cognitive/learning and social/personality factors. Applied Cognitive Psychology, 25(4), 528–535. https://doi.org/10.1002/acp.1725</bibtext> </blist> <blist> <bibtext> Henson R., Templin J., Willse J. (2009). Defining a family of cognitive diagnosis models using log-linear models with latent variables. Psychometrika, 74(2), 191–210. https://doi.org/10.1007/s11336-008-9089-5</bibtext> </blist> <blist> <bibtext> Hou L., de la Torre J., Nandakumar R. (2014). Differential item functioning assessment in cognitive diagnostic modeling: Application of the Wald test to investigate DIF in the DINA model. Journal of Educational Measurement, 51(1), 98–125. https://doi.org/10.1111/jedm.12036</bibtext> </blist> <blist> <bibtext> Huebner A., Wang C. (2011). A note on comparing examinee classification methods for cognitive diagnosis models. Educational and Psychological Measurement, 71(2), 407–419. https://doi.org/10.1177/0013164410388832</bibtext> </blist> <blist> <bibtext> Iaconangelo C., de la Torre J. (2016). Three-step estimation of cognitive diagnosis models with covariate. In Paper presented at the international meeting of the psychometric society, Asheville, NC, USA.</bibtext> </blist> <blist> <bibtext> Jeffreys H. (1961). Theory of probability (3rd ed.). Oxford University Press.</bibtext> </blist> <blist> <bibtext> Junker B. W., Sijtsma K. (2001). Cognitive assessment models with few assumptions, and connections with nonparametric item response theory. Applied Psychological Measurement, 25(3), 258–272. https://doi.org/10.1177/01466210122032064</bibtext> </blist> <blist> <bibtext> Koponen T., Eklund K. M., Heikkilä R., Salminen J., Fuchs L. S., Fuchs D., Aro M. (2020). Cognitive correlates of the covariance in reading and arithmetic fluency: Importance of serial retrieval fluency. Child Development, 91(4), 1063–1080. https://doi.org/10.1111/cdev.13287</bibtext> </blist> <blist> <bibtext> Kreitchmann R. S., de la Torre J., Sorrel M. A., Najera P., Abad F. J. (2022). Improving reliability estimation in cognitive diagnosis modeling. Behavior Research Methods, 55, 3446–3460. https://doi.org/10.3758/s13428-022-01967-5</bibtext> </blist> <blist> <bibtext> Lee S.-J., Chae Y.-G. (2007). Children's internet use in a family context: Influence on family relationships and parental mediation. CyberPsychology and Behavior, 10(5), 640–644. https://doi.org/10.1089/cpb.2007.9975</bibtext> </blist> <blist> <bibtext> Lee Y. S., Park Y. S., Taylan D. (2011). A cognitive diagnostic modeling of attribute mastery in Massachusetts, Minnesota, and the US national sample using the TIMSS 2007. International Journal of Testing, 11(2), 144–177. https://doi.org/10.1080/15305058.2010.534571</bibtext> </blist> <blist> <bibtext> Liang Q., de la Torre J., Law N. (2021). Do background characteristics matter in children's mastery of digital literacy? A cognitive diagnosis model analysis. Computers in Human Behavior, 122, 106850. https://doi.org/10.1016/j.chb.2021.106850</bibtext> </blist> <blist> <bibtext> Ma W., de la Torre J. (2020). Gdina: The generalized DINA model framework. Comprehensive R Archive Network (CRAN). <ulink href="http://CRAN.R-project.org/package=GDINA">http://CRAN.R-project.org/package=GDINA</ulink></bibtext> </blist> <blist> <bibtext> Ma W., Terzi R., de la Torre J. (2021). Detecting differential item functioning using multiple-group cognitive diagnosis models. Applied Psychological Measurement, 45(1), 37–53. https://doi.org/10.1177/0146621620965745</bibtext> </blist> <blist> <bibtext> Magis D., Beland S., Tuerlinckx F., de Boeck P. (2010). A general framework and an R package for the detection of dichotomous differential item functioning. Behavior Research Methods, 42(3), 847–862. https://doi.org/10.3758/BRM.42.3.847</bibtext> </blist> <blist> <bibtext> Maris E. (1999). Estimating multiple classification latent class models. Psychometrika, 64(2), 187–212. https://doi.org/10.1007/BF0229453</bibtext> </blist> <blist> <bibtext> American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. American Educational Research Association.</bibtext> </blist> <blist> <bibtext> Rothstein J. M. (2004). College performance predictions and the SAT. Journal of Econometrics, 121, 297–317. https://doi.org/10.1016/j.jeconom.2003.10.003</bibtext> </blist> <blist> <bibtext> Rupp A. A., Templin J. L., Henson R. A. (2010). Diagnostic measurement: Theory, methods, and applications. The Guilford Press.</bibtext> </blist> <blist> <bibtext> Sackett P. R., Kuncel N. R., Beatty A. S., Rigdon J. L., Shen W., Kiger T. B. (2012). The role of socioeconomic status in sat-grade relationships and in college admissions decisions. Psychological Science, 23(9), 1000–1007. https://doi.org/10.1177/0956797612438732</bibtext> </blist> <blist> <bibtext> Salthouse T. A. (2019). Trajectories of normal cognitive aging. Psychology and Aging, 34(1), 17–24. https://doi.org/10.1037/pag0000288</bibtext> </blist> <blist> <bibtext> Schmitt N., Oswald F. L., Kim B. H., Imus A., Merritt S., Friede A., Shivpuri S. (2007). The use of background and ability profiles to predict college student outcomes. Journal of Applied Psychology, 92(1), 165–179. https://doi.org/10.1037/0021-9010.92.1.165</bibtext> </blist> <blist> <bibtext> Stage C. W., Ögren G. (2004). The Swedish Scholastic Assessment Test (SweSAT): Development, results and experiences. Umea University.</bibtext> </blist> <blist> <bibtext> Sun X., Wang S., Guo L., Xin T., Song N. (2023). Using a generalized logistic regression method to detect differential item functioning with multiple groups in cognitive diagnostic tests. Applied Psychological Measurement, 47(4), 328–346. https://doi.org/10.1177/01466216231174559</bibtext> </blist> <blist> <bibtext> Tatsuoka K. K. (1983). Rule space: An approach for dealing with misconceptions based on item response theory. Journal of Educational Measurement, 20(4), 345–354. https://doi.org/10.1111/j.1745-3984.1983.tb00212.x</bibtext> </blist> <blist> <bibtext> Templin J. L., Henson R. A. (2006). Measurement of psychological disorders using cognitive diagnosis models. Psychological Methods, 11(3), 287. https://doi.org/10.1037/1082-989X.11.3.287</bibtext> </blist> <blist> <bibtext> Toprak-Yildiz T. E. (2021). An international comparison using cognitive diagnostic assessment: Fourth graders' diagnostic profile of reading skills on PIRLS 2016. Studies in Educational Evaluation, 70, 101057. https://doi.org/10.1016/j.stueduc.2021.101057</bibtext> </blist> <blist> <bibtext> Vermunt J. K. (2010). Latent class modeling with covariates: Two improved three-step approaches. Political Analysis, 18(4), 450–469. https://doi.org/10.1093/pan/mpq025</bibtext> </blist> <blist> <bibtext> von Davier M. (2008). A general diagnostic model applied to language testing data. British Journal of Mathematical and Statistical Psychology, 61(2), 287–307. https://doi.org/10.1348/000711007X193957</bibtext> </blist> <blist> <bibtext> von Davier M., Lee Y-S. (2019). Handbook of diagnostic classification models (1st ed.). Springer, https://doi.org/10.1007/978-3-030-05584-4</bibtext> </blist> <blist> <bibtext> Wang J., Duan J. (2024). Determining the number of attributes in the GDINA model. British Journal of Mathematical and Statistical Psychology, 78(1), 1–28. https://doi.org/10.1111/bmsp.12349</bibtext> </blist> <blist> <bibtext> Wedman J. (2017). Reasons for gender-related differential item functioning in a college admissions test. Scandinavian Journal of Educational Research, 62(6), 959–970. https://doi.org/10.1080/00313831.2017.1402365</bibtext> </blist> <blist> <bibtext> Wu X., Wu R., Chang H. H., Kong Q., Zhang Y. (2020). International comparative study on PISA mathematics achievement test based on cognitive diagnostic models. Frontiers in Psychology, 11, 2230. https://doi.org/10.3389/fpsyg.2020.02230</bibtext> </blist> <blist> <bibtext> Xu T., Wu X., Sun S., Kong Q. (2023). Cognitive diagnostic analysis of students' mathematical competency based on the DINA model. Psychology in the Schools, 20(9), 3135–3150. https://doi.org/10.1002/pits.22916</bibtext> </blist> <blist> <bibtext> Zwick R., Green J. G. (2007). New perspectives on the correlation of SAT scores, high school grades, and socioeconomic factors. Journal of Educational Measurement, 44(1), 23–45. https://doi.org/10.1111/j.1745-3984.2007.00025.x</bibtext> </blist> </ref> <ref id="AN0193364109-28"> <title> Footnotes </title> <blist> <bibtext> Könül Karimova https://orcid.org/0000-0003-4840-3573 Inga Laukaityte https://orcid.org/0000-0001-7282-5384 Christina Wikström https://orcid.org/0000-0002-4625-4853</bibtext> </blist> <blist> <bibtext> Könül Karimova conceived and designed the analysis, performed the analysis, and wrote the paper. Inga Laukaityte contributed with the statistical analysis, reviewed and revised the manuscript. Christina Wikström provided the data, reviewed, revised, and added information about SweSAT in the manuscript.</bibtext> </blist> <blist> <bibtext> The authors received no financial support for the research, authorship, and/or publication of this article.</bibtext> </blist> <blist> <bibtext> The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.</bibtext> </blist> </ref> <aug> <p>By Könül Karimova; Inga Laukaityte and Christina Wikström</p> <p>Reported by Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib14" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib31" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib35" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib48" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib21" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib30" firstref="ref10"></nolink> <nolink nlid="nl7" bibid="bib36" firstref="ref14"></nolink> <nolink nlid="nl8" bibid="bib33" firstref="ref15"></nolink> <nolink nlid="nl9" bibid="bib23" firstref="ref16"></nolink> <nolink nlid="nl10" bibid="bib47" firstref="ref17"></nolink> <nolink nlid="nl11" bibid="bib46" firstref="ref18"></nolink> <nolink nlid="nl12" bibid="bib40" firstref="ref19"></nolink> <nolink nlid="nl13" bibid="bib25" firstref="ref20"></nolink> <nolink nlid="nl14" bibid="bib27" firstref="ref22"></nolink> <nolink nlid="nl15" bibid="bib13" firstref="ref29"></nolink> <nolink nlid="nl16" bibid="bib45" firstref="ref30"></nolink> <nolink nlid="nl17" bibid="bib41" firstref="ref33"></nolink> <nolink nlid="nl18" bibid="bib24" firstref="ref34"></nolink> <nolink nlid="nl19" bibid="bib18" firstref="ref42"></nolink> <nolink nlid="nl20" bibid="bib15" firstref="ref46"></nolink> <nolink nlid="nl21" bibid="bib20" firstref="ref47"></nolink> <nolink nlid="nl22" bibid="bib32" firstref="ref48"></nolink> <nolink nlid="nl23" bibid="bib42" firstref="ref49"></nolink> <nolink nlid="nl24" bibid="bib43" firstref="ref50"></nolink> <nolink nlid="nl25" bibid="bib38" firstref="ref52"></nolink> <nolink nlid="nl26" bibid="bib29" firstref="ref53"></nolink> <nolink nlid="nl27" bibid="bib39" firstref="ref55"></nolink> <nolink nlid="nl28" bibid="bib28" firstref="ref61"></nolink> <nolink nlid="nl29" bibid="bib16" firstref="ref62"></nolink> <nolink nlid="nl30" bibid="bib22" firstref="ref65"></nolink> <nolink nlid="nl31" bibid="bib10" firstref="ref67"></nolink> <nolink nlid="nl32" bibid="bib19" firstref="ref70"></nolink> <nolink nlid="nl33" bibid="bib37" firstref="ref71"></nolink> <nolink nlid="nl34" bibid="bib26" firstref="ref77"></nolink> <nolink nlid="nl35" bibid="bib17" firstref="ref78"></nolink> <nolink nlid="nl36" bibid="bib44" firstref="ref80"></nolink> <nolink nlid="nl37" bibid="bib12" firstref="ref81"></nolink> <nolink nlid="nl38" bibid="bib11" firstref="ref85"></nolink> <nolink nlid="nl39" bibid="bib34" firstref="ref87"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1504546
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: The Cognitive Diagnostic Model Analysis of the Relationship between Verbal and Quantitative Skills with Background Characteristics: Evidence from the Swedish Scholastic Assessment Test 2023
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Könül+Karimova%22">Könül Karimova</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-4840-3573">0000-0003-4840-3573</externalLink>)<br /><searchLink fieldCode="AR" term="%22Inga+Laukaityte%22">Inga Laukaityte</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-7282-5384">0000-0001-7282-5384</externalLink>)<br /><searchLink fieldCode="AR" term="%22Christina+Wikström%22">Christina Wikström</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-4625-4853">0000-0002-4625-4853</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Advanced+Academics%22"><i>Journal of Advanced Academics</i></searchLink>. 2026 37(2):364-392.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 29
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2026
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22High+Schools%22">High Schools</searchLink><br /><searchLink fieldCode="EL" term="%22Secondary+Education%22">Secondary Education</searchLink><br /><searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Aptitude+Tests%22">Aptitude Tests</searchLink><br /><searchLink fieldCode="DE" term="%22High+Stakes+Tests%22">High Stakes Tests</searchLink><br /><searchLink fieldCode="DE" term="%22Verbal+Ability%22">Verbal Ability</searchLink><br /><searchLink fieldCode="DE" term="%22Mathematics+Skills%22">Mathematics Skills</searchLink><br /><searchLink fieldCode="DE" term="%22Cognitive+Tests%22">Cognitive Tests</searchLink><br /><searchLink fieldCode="DE" term="%22Models%22">Models</searchLink><br /><searchLink fieldCode="DE" term="%22Test+Wiseness%22">Test Wiseness</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Characteristics%22">Student Characteristics</searchLink><br /><searchLink fieldCode="DE" term="%22Measurement%22">Measurement</searchLink><br /><searchLink fieldCode="DE" term="%22Gender+Differences%22">Gender Differences</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Attainment%22">Educational Attainment</searchLink><br /><searchLink fieldCode="DE" term="%22Age+Differences%22">Age Differences</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Evaluation%22">Student Evaluation</searchLink><br /><searchLink fieldCode="DE" term="%22Item+Response+Theory%22">Item Response Theory</searchLink><br /><searchLink fieldCode="DE" term="%22Multiple+Choice+Tests%22">Multiple Choice Tests</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Tests%22">Reading Tests</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Comprehension%22">Reading Comprehension</searchLink><br /><searchLink fieldCode="DE" term="%22College+Students%22">College Students</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Sweden%22">Sweden</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1177/1932202X251371339
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1932-202X<br />2162-9536
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: For decades, researchers have sought analytic methods that yield diagnostic information about test-takers while ensuring that high-stakes tests remain free of construct-irrelevant bias. This study applies a general cognitive-diagnostic-model (CDM) framework to analyze the quantitative and verbal skills (QVS) assessed by the Swedish Scholastic Aptitude Test (SweSAT). A three-step latent-class logistic-regression approach was used to investigate the relationship between test-takers' background characteristics and performance in each domain subskill. The analysis was conducted using representative data from 41,451 test-takers from the 2023 administration of the SweSAT, focusing on performance across four quantitative and four verbal subskills, and examining test characteristics. The results showed that the CDM method was appropriate for analyzing QVS, with evidence of measurement invariance across sex, age, and educational levels in the subskills. Additionally, the findings revealed distinct associations between sex, educational level, and age with performance in each QVS subskill. Implications for equitable selection in higher education are discussed.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2026
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1504546
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1504546
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1177/1932202X251371339
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 29
        StartPage: 364
    Subjects:
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Aptitude Tests
        Type: general
      – SubjectFull: High Stakes Tests
        Type: general
      – SubjectFull: Verbal Ability
        Type: general
      – SubjectFull: Mathematics Skills
        Type: general
      – SubjectFull: Cognitive Tests
        Type: general
      – SubjectFull: Models
        Type: general
      – SubjectFull: Test Wiseness
        Type: general
      – SubjectFull: Student Characteristics
        Type: general
      – SubjectFull: Measurement
        Type: general
      – SubjectFull: Gender Differences
        Type: general
      – SubjectFull: Educational Attainment
        Type: general
      – SubjectFull: Age Differences
        Type: general
      – SubjectFull: Student Evaluation
        Type: general
      – SubjectFull: Item Response Theory
        Type: general
      – SubjectFull: Multiple Choice Tests
        Type: general
      – SubjectFull: Reading Tests
        Type: general
      – SubjectFull: Reading Comprehension
        Type: general
      – SubjectFull: College Students
        Type: general
      – SubjectFull: Sweden
        Type: general
    Titles:
      – TitleFull: The Cognitive Diagnostic Model Analysis of the Relationship between Verbal and Quantitative Skills with Background Characteristics: Evidence from the Swedish Scholastic Assessment Test 2023
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Könül Karimova
      – PersonEntity:
          Name:
            NameFull: Inga Laukaityte
      – PersonEntity:
          Name:
            NameFull: Christina Wikström
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 1932-202X
            – Type: issn-electronic
              Value: 2162-9536
          Numbering:
            – Type: volume
              Value: 37
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Journal of Advanced Academics
              Type: main
ResultId 1