Uniform Differential Item Functioning Research Articles

Examining Differential Item Functioning from a Multidimensional IRT Perspective.

Differential item functioning (DIF) is a standard analysis for every testing company. Research has demonstrated that DIF can result when test items measure different ability composites, and the groups being examined for DIF exhibit distinct underlying ability distributions on those composite abilities. In this article, we examine DIF from a two-dimensional multidimensional item response theory (MIRT) perspective. We begin by delving into the compensatory MIRT model, illustrating and how items and the composites they measure can be graphically represented. Additionally, we discuss how estimated item parameters can vary based on the underlying latent ability distributions of the examinees. Analytical research highlighting the consequences of ignoring dimensionally and applying unidimensional IRT models, where the two-dimensional latent space is mapped onto a unidimensional, is reviewed. Next, we investigate three different approaches to understanding DIF from a MIRT standpoint: 1. Analytically Uniform and Nonuniform DIF: When two groups of interest have different two-dimensional ability distributions, a unidimensional model is estimated. 2. Accounting for complete latent ability space: We emphasize the importance of considering the entire latent ability space when using DIF conditional approaches, which leads to the mitigation of DIF effects. 3. Scenario-Based DIF: Even when underlying two-dimensional distributions are identical for two groups, differing problem-solving approaches can still lead to DIF. Modern software programs facilitate routine DIF procedures for comparing response data from two identified groups of interest. The real challenge is to identify why DIF could occur with flagged items. Thus, as a closing challenge, we present four items (Appendix A) from a standardized test and invite readers to identify which group was favored by a DIF analysis.

Sex-related differential item functioning of the Jenkins Sleep Scale: a cross-sectional study among 77 967 employees in the Finnish public sector

ObjectivesTo investigate if the Jenkins Sleep Scale (JSS) demonstrates sex-related differential item functioning (DIF).DesignCross-sectional study.SettingSurvey data from the Finnish Public Sector study (2015–2017).Participants77 967 employees in the Finnish public sector,...

Detecting uniform differential item functioning for continuous response computerized adaptive testing

Evaluating items for potential differential item functioning (DIF) is an essential step to ensuring measurement fairness. In this article, we focus on a specific scenario, namely, the continuous response, severely sparse, computerized adaptive testing (CAT). Continuous responses items are growingly used in performance-based tasks because they tend to generate more information than traditional dichotomous items. Severe sparsity arises when many items are automatically generated via machine learning algorithms. We propose two uniform DIF detection methods in this scenario. The first is a modified version of the CAT-SIBTEST, a non-parametric method that does not depend on any specific item response theory model assumptions. The second is a regularization method, a parametric, model-based approach. Simulation studies show that both methods are effective in correctly identifying items with uniform DIF. A real data analysis is provided in the end to illustrate the utility and potential caveats of the two methods.

Measurement equivalence of the English and French versions of the self-efficacy to manage chronic disease scale: a Scleroderma Patient-Centered Intervention Network (SPIN) study.

The Self-Efficacy to Manage Chronic Disease (SEMCD) scale is widely used, including in systemic sclerosis (SSc). The SEMCD has been validated in SSc, but the metric equivalence of the English and French versions has not been assessed (i.e., whether psychometric properties are equivalent across English and French). Participants were adults from the Scleroderma Patient-Centered Intervention Network (SPIN) Cohort (N = 2159) who completed baseline measures in English (n = 1473) or French (n = 686) between May 2014 to July 2020. Analyses assessed internal consistency reliability via Cronbach's alpha and McDonald's omega, convergent validity via Pearson's correlations, structural validity via confirmatory factor analysis (CFA), and differential item functioning via the Multiple-Indicator Multiple-Cause (MIMIC) model. Internal consistency reliability was high in English (α = .93, ω = .93) and French (α = .92, ω = .93). All correlations between the SEMCD and measures of health outcomes were moderate to large, statistically significant, and in the hypothesized direction in both languages. The CFA demonstrated that the one-factor model of self-efficacy, overall, fit reasonably well (CFI = .96, TLI = .93, SRMR = .03, RMSEA = .14). Standardized factor loadings were large (.76 to .88). Three items displayed statistically significant uniform DIF and all six displayed nonuniform DIF; all DIF was of minimal magnitude. Comparison of unadjusted and DIF-adjusted models indicated that DIF did not meaningfully impact total score (ICC = 0.999, r = 0.999). Scores from English- and French-speaking adults with SSc can be combined for analysis or compared.

FURTHER EVALUATION OF A MODIFIED VERSION OF THE HOSPICE PHILOSOPHY SCALE: A DIFFERENTIAL ITEM FUNCTIONING ANALYSIS

Abstract The hospice philosophy of care (HPOC) is a set of values applicable to all interdisciplinary group members that drives organizational and individual approaches to hospice care. As the only established measure of HPOC attitudes, we aimed to extend previous exploratory psychometric evaluation of the eight-item Hospice Philosophy Scale (HPS-8). An interdisciplinary convenience sample of 481 hospice clinicians participated in this cross-sectional study. Participants were recruited through hospice and palliative care membership associations representing the core members of the Medicare Hospice Benefit-designated hospice interdisciplinary group (physicians, nurses, social workers, chaplains). After examining uniformity through confirmatory factor analysis, we tested HPS-8 items for differential item functioning (DIF) by professional discipline using multiple indicators, multiple causes models. We compared models via nested chi-square difference tests. Our final model assessed uniform DIF for the HPS-8 factor, controlling for DIF-indicated items. We also assessed internal consistency reliability. After correlating errors for two similarly worded items, global fit indices met prevailing thresholds, χ²(19)=55.18, p<.001 (RMSEA=.06, SRMR=.03, CFI=.98, TLI=.97). The uniform DIF model indicated two items with DIF. Chaplains endorsed the need for mental and spiritual preparation for death more than physicians (β=0.21, p<.001). Nurses endorsed involving patients and families in deciding treatments and services more than physicians (β=0.21, p=.006). Despite statistically significant DIF on these items, statistically significant differences among professions were not observed at the factor level (p>.50 for all). Composite reliability estimates satisfied internal consistency thresholds (CR=.76). Results support the HPS-8’s use as a valid and reliable measure for assessing HPOC in hospice clinicians.

30 Item response theory and differential item functioning of the AD8: The High School & Beyond Study

Objective:The AD8 is a validated screening instrument for functional changes that may be caused by cognitive decline and dementia. It is frequently used in clinics and research studies because it is short and easy to administer, with a cut off score of 2 out of 8 items recommended to maximize sensitivity and specificity. This cutoff assumes that all 8 items provide equivalent “information” about everyday functioning. In this study, we used item response theory (IRT) to test this assumption. To determine the relevance of this measure of everyday functioning in men and women, and across race, ethnicity, and education, we conducted differential item functioning (DIF) analysis to test for item bias.Participants and Methods:Data came from the 2021 follow up of the High School & Beyond cohort (N=8,690; mean age 57.5 ± 1.2; 55% women), a nationally representative, longitudinal study of Americans who were first surveyed in 1980 when they were in the 10th or 12th grade. Participants were asked AD8 questions about their own functioning via phone or internet survey. First, we estimated a one-parameter (i.e., differing difficulty, equal discrimination across items) and two-parameter IRT model (i.e., differing difficulty and differing discrimination across items). We compared model fit using a likelihood-ratio test. Second, we tested for uniform and non-uniform DIF on AD8 items by sex, race and ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic), education level (high school or less, some college, BA degree or more), and survey mode (phone or internet). We examined DIF salience by comparing the difference between original and DIF-adjusted AD8 scores to the standard error of measurement of the original score.Results:The two-parameter IRT model fit the data significantly better than the one-parameter model, indicating that some items were more strongly related to underlying everyday functional ability than others. For example, the “problems with judgment” item had higher discrimination (more information) than the “less interest in hobbies/activities” item. There were significant differences in item endorsement by race/ethnicity, education, and survey mode. We found significant uniform and non-uniform DIF on several items across each of these groups. For example, for a given level of functional decline (theta) White participants were more likely to endorse “Daily problems with thinking/memory” than Black and Hispanic participants. The DIF was salient (i.e., caused AD8 scores to change by greater than the standard error of measurement for a large portion of respondents) for those with a college degree and phone respondents.Conclusions:In a population representative sample of Americans ∼age 57, the items on the AD8 contributed differing levels of discrimination along the range of everyday functioning that is impacted by later life cognitive impairment. This suggests that a simple cut-off or summed score may not be appropriate since some items yield more information about the underlying construct than others. Furthermore, we observed significant and salient DIF on several items by education and survey mode, AD8 scores should not be compared across education groups and assessment modes without adjustment for this measurement bias.

The Investigation of The Differential Item Functions of The 2012 High School Entrance Exam Mathematics Test Based On The G-DINA Model

Purpose: This study aims to determine the item parameters and estimations of student attribute profiles related to the 2012 8th Grade HSEE Mathematics subtest item responses using the G-DINA model and examine whether the items show DIF according to the gender variable.
 Design/Methodology/Approach: The study was conducted on the data of 1.063.570 students who took the 2012 HSEE Mathematics subtest. The analysis was carried out on the target population data to avoid sampling error. This study is a descriptive, survey type study.
 Findings: When the model fit indices and comparison results were examined, it was concluded that the model that best explained the behavior of students responding to the items in the 20-item form of the 2012 HSEE Mathematics subtest was G- DINA. When the G-DINA model parameters were examined, it was found that the 𝛼! = [0000] attribute profile (with 61%) in which none of the four defined attributes were found in the student, and the 𝛼!" = [1111] attribute profile (with 17%) in which all four attributes were present in the student (17%) were the most common attribute profiles. As a result of DIF analysis within the scope of CDM, and it was identified that item 4 showed a significant uniform DIF in favor of female students while item 19 showed a uniform DIF in favor of male students at moderate level.
 Highlights: It is thought that DIF analyses within the framework of cognitive diagnostic models can provide a statistical basis for item bias decisions.

A Comparison of Mantel-Haenszel and the Area Difference Between Item Characteristic Curves Index Techniques for Detecting Differential Item Functioning

The study is aimed at comparing between Mantel-Haenszel-MH chi-square and the area difference between three item characteristic curves Index -3ICCs by using Raju’s index to detect the differential item functioning DIF in post-graduate general aptitude test according to gender variable. The study sample of (2280) students was distributed equally between male and female groups. The MH techniques revealed a uniform DIF in (43) items, equal to (45%) of the items in post-graduate general aptitude test. 3ICCs technique, revealed a uniform DIF based on gender in two items, equals to (2%) of item's test. MH technique revealed more efficiency in detecting differential item functioning in post-graduate general aptitude test compared to area difference between 3ICCs. The agreement rate was low between these two techniques. Based on these results, the researchers recommended to use more than one technique when detecting the differential item functioning. Keywords: Differential Item Functioning, Mantel-Haenszel, The three-parameter logistic model, post-graduate general aptitude test.

Measuring Gelotophobia, Gelotophilia, and Katagelasticism in Italy and Canada Using PhoPhiKat-30

Abstract: The PhoPhiKat-30 is a self-report instrument for describing personality related to laughter and ridicule including gelotophobia, gelotophilia, and katagelasticism. The present study assessed the measurement properties of the newly translated Italian PhoPhiKat-30 across participants in Italy and Canada using multidimensional item response theory. Italian ( N = 326) and Canadian ( N = 1,467) participants completed the Italian and English PhoPhiKat-30, respectively. The parallel analysis supported the three-factor model in Italy. Conditional reliability estimates showed strong precision (> 0.80) of gelotophobia and gelotophilia along the latent continuum (−1.15 < θ < 3.08 and −1.69 < θ < 3.09, respectively). Katagelasticism showed a limited range (0.98 < θ < 2.85) for the latent attribute precisely measured, suggesting that new items that address the low to moderate difficulty of katagelasticism should be added in future studies. Item discrimination parameters varied across Reckase’s multidimensional normal-ogive model (MDISC mean = 0.79). Five items had uniform differential item functioning (DIF; McFadden’s pseudo R2 > .035 or β > .10) when comparing the Italian and English PhoPhiKat-30, with English items showing more agreement at the same level of the latent trait. The Italian PhoPhiKat-30 has good item discrimination across the latent continuum and showed cross-cultural equivalence for most items.

Comparing different trend estimation approaches in country means and standard deviations in international large-scale assessment studies

One major aim of international large-scale assessments (ILSA) like PISA is to monitor changes in student performance over time. To accomplish this task, a set of common items (i.e., link items) is repeatedly administered in each assessment. Linking methods based on item response theory (IRT) models are used to align the results from the different assessments on a common scale. This work employs the one-parameter logistic (1PL) and the two-parameter logistic (2PL) IRT models as scaling models for dichotomous item response data. The present article discusses different types of trend estimates in country means and standard deviations for countries in ILSA. These types differ in three aspects. First, the trend can be assessed by an indirect or direct linking approach for linking a country’s performance at an international metric. Second, the linking for the trend estimation can rely on either all items or only the link items. Third, item parameters can be assumed to be invariant or noninvariant across countries. It is shown that the most often employed trend estimation methods of original trends and marginal trends can be conceived as particular cases of indirect and direct linking approaches, respectively. Through a simulation study and analytical derivations, it is demonstrated that trend estimates using a direct linking approach and those that rely on only link items outperformed alternatives for the 1PL model with uniform country differential item functioning (DIF) and the 2PL model with uniform and nonuniform country DIF. We also illustrated the performance of the different scaling models for assessing the PISA trend from PISA 2006 to PISA 2009 in the cognitive domains of reading, mathematics, and science. In this empirical application, linking errors based on jackknifing testlets were utilized that adequately quantify DIF effects in the uncertainty of trend estimates.

Evaluation of psychometric properties of the Dental Anxiety Inventory (DAI-36) questionnaire using iterative hybrid ordinal logistic: Differential item functioning (DIF).

Dental Anxiety Inventory (DAI-36) questionnaire is an instrument for assessing dental anxiety. The different perceptions of the questionnaire items in the individual at the same level of underlying dental anxiety may lead to different reported dental anxieties. This study aims to determine the differential item functioning (DIF) of the DAI-36 measure items. The DAI-36 was completed by 950 participants. An iterative hybrid ordinal logistic regression model was used to detect DIF across gender, education, and age groups. DIF analysis was done by lordif package in R3.1.3 software. The chi-square statistics declared 7, 7, and 4 nonuniform DIF items, and 2, 5, and 4 uniform DIF items across gender, education, and age groups, respectively. ΔR2 was always lower than 0.07 in all uniform and nonuniform DIF items. However, Δβ1 revealed significant uniform DIF in items 1 and 8 across gender (Δβ1(item 1)=0.0137, Δβ1(item 8)=0.0124) and items 22 and 27 across age categories (Δβ1(item 22)=0.0110, Δβ1(item 27)=0.0136). DIF items had no large magnitude or cancel out each other, so statements phrased in the DAI-36 questionnaire have equivalent meaning across participants, regardless of their gender, education, and age groups.

Wald χ2 Test for Differential Item Functioning Detection with Polytomous Items in Multilevel Data

Identifying items with differential item functioning (DIF) in an assessment is a crucial step for achieving equitable measurement. One critical issue that has not been fully addressed with existing studies is how DIF items can be detected when data are multilevel. In the present study, we introduced a Lord’s Wald [Formula: see text] test-based procedure for detecting both uniform and non-uniform DIF with polytomous items in the presence of the ubiquitous multilevel data structure. The proposed approach is a multilevel extension of a two-stage procedure, which identifies anchor items in its first stage and formally evaluates candidate items in the second stage. We applied the Metropolis–Hastings Robbins–Monro (MH-RM) algorithm to estimate multilevel polytomous item response theory (IRT) models and to obtain accurate covariance matrices. To evaluate the performance of the proposed approach, we conducted a preliminary simulation study that considered various conditions to mimic real-world scenarios. The simulation results indicated that the proposed approach has great power for identifying DIF items and well controls the Type I error rate. Limitations and future research directions were also discussed.

Using Rasch Analysis to Assess and Improve the Measurement Properties of a Questionnaire With Few Items: The York Binaural Hearing-Related Quality of Life (YBHRQL) Questionnaire.

The York Binaural Hearing-Related Quality of Life questionnaire is a condition-specific preference-based instrument sensitive to advantages of binaural over monaural hearing. Respondents use 5-point scales to report the difficulty experienced with three dimensions of listening that are easier or more successful when hearing is binaural: understanding speech in spatially separated noise, localizing sources of sound in azimuth, and the associated effort and fatigue. Previously, a preference value was estimated for each combination of dimension and level so that a value of binaural utility could be assigned to a respondent to inform analyses of cost effectiveness. The present objective was to determine whether the questionnaire conforms with the Rasch model sufficiently well for estimates of the binaural abilities of respondents to be obtained on an interval scale to inform parametric analyses of clinical effectiveness. Data were obtained from unilateral cochlear implantees (N = 418; 209 ≤62 years; 209 ≥63 years) and members of the public (N = 325; 207 ≤62 years; 118 ≥63 years). A subset of implantees (N = 118) responded at test and retest. Responses were fitted to the partial credit model using the Extended Rasch Modeling package. Conformity with the model was evaluated in six ways: the ordering of response categories ( Monotonicity ) was assessed with plots of response probability against ability; differential item functioning ( DIF ) was assessed by analyses of variance of standardized response residuals; alignment of participants' abilities with item difficulties ( Targeting ) was assessed with person-item maps; fit to the model ( Fit ) was assessed by comparing the means and variabilities of observed and expected responses, and by comparing observed values with analyses of simulated datasets; the hypothesis that item difficulties and participants' abilities were measured on a single underlying scale ( Unidimensionality ) was assessed with principal components analyses of standardized response residuals. Values of fit statistics were toward the lower end of the acceptable range. Comparisons with analyses of simulated datasets showed that low values were primarily the result of the structural limitation of including only three items. Modal values of the probabilities of response categories were ordered monotonically, but some response thresholds were disordered because of under-use of one category. Pooling categories to correct disordered thresholds resulted in estimates of ability that were less discriminatory of differences within and between groups, and showed less reproducibility between test and retest, than did the original estimates. Neither source-related DIF nor gender-related DIF arose. Uniform age-related DIF arose for the speech-in-noise item and could be managed by resolving the item. The resulting estimates of ability and difficulty were well targeted and unidimensional. The York Binaural Hearing-Related Quality of Life questionnaire, with three items each with five response categories, conforms with the Rasch model sufficiently well to yield practically useful measures of the abilities of participants. The trait measured by the questionnaire aligns with the ability to benefit from binaural hearing. More discriminatory measurement of this ability would be achieved with more items. Nonetheless, the questionnaire possesses the virtue that responses to the same three questions can be scored in different ways to inform parametric analyses of both cost-effectiveness and clinical effectiveness.

Detecting Differential Item Functioning in 2PL Multistage Assessments

The detection of differential item functioning is crucial for the psychometric evaluation of multistage tests. This paper discusses five approaches presented in the literature: logistic regression, SIBTEST, analytical score-based tests, bootstrap score-based tests, and permutation score-based tests. First, using an simulation study inspired by a real-life large-scale educational assessment, we compare the five approaches with respect to their type I error rate and their statistical power. Then, we present an application to an empirical data set. We find that all approaches show type I error rates close to the nominal alpha level. Furthermore, all approaches are shown to be sensitive to uniform and non-uniform DIF effects, with the score-based tests showing the highest power.

Detecting Differential Item Functioning in CAT Using IRT Residual DIF Approach

AbstractThe residual differential item functioning (RDIF) detection framework was developed recently under a linear testing context. To explore the potential application of this framework to computerized adaptive testing (CAT), the present study investigated the utility of the RDIFR statistic both as an index for detecting uniform DIF of pretest items in CAT and as a direct measure of the effect size of uniform DIF. Extensive CAT simulations revealed RDIFR to have well‐controlled Type I error and slightly higher power to detect uniform DIF compared with CATSIB, especially when pretest items were calibrated using fixed‐item parameter calibration. Moreover, RDIFR accurately estimated the amount of uniform DIF irrespective of the presence of impact. Therefore, RDIFR demonstrates its potential as a useful tool for evaluating both the statistical and practical significance of uniform DIF in CAT.

Multi-Group Generalizations of SIBTEST and Crossing-SIBTEST

ABSTRACT This article presents generalizations of SIBTEST and crossing-SIBTEST statistics for differential item functioning (DIF) investigations involving more than two groups. After reviewing the original two-group setup for these statistics, a set of multigroup generalizations that support contrast matrices for joint tests of DIF are presented. To investigate the Type I error and power behavior of these generalizations, a Monte Carlo simulation study was then explored. Results indicated that the proposed generalizations are reasonably effective at recovering their respective population parameter definitions, maintain optimal Type I error control, have suitable power to detect uniform and non-uniform DIF, and in shorter tests are competitive with the generalized logistic regression and generalized Mantel–Haenszel tests for DIF.

Do the SCOFF items function differently by food-security status in U.S. college students?: Statistically, but not practically, significant differences

Despite food insecurity (FI) being associated with eating disorders (EDs), little research has examined if ED screening measures perform differently in individuals with FI. This study tested whether items on the SCOFF performed differently as a function of FI. As many people with FI hold multiple marginalized identities, this study also tested if the SCOFF performs differently as a function of food-security status in individuals with different gender identities and different perceived weight statuses. Data were from the 2020/2021 Healthy Minds Study (N = 122,269). Past-year FI was established using the two-item Hunger Vital Sign. Differential item functioning (DIF) assessed whether SCOFF items performed differently (i.e., had different probabilities of endorsement) in groups of individuals with FI versus those without. Both uniform DIF (constant between-group difference in item-endorsement probability across ED pathology) and non-uniform DIF (variable between-group difference in item-endorsement probability across ED pathology) were examined. Several SCOFF items demonstrated both statistically significant uniform and non-uniform DIF (ps < .001), but no instances of DIF reached practical significance (as indicated by effect sizes pseudo ΔR2 ≥ 0.035; all pseudo ΔR2's ≤ 0.006). When stratifying by gender identity and weight status, although most items demonstrated statistically significant DIF, only the SCOFF item measuring body-size perception showed practically significant non-uniform DIF for perceived weight status. Findings suggest the SCOFF is an appropriate screening measure for ED pathology among college students with FI and provide preliminary support for using the SCOFF in individuals with FI and certain marginalized identities.

A Comparison of the efficacies of differential item functioning detection methods

To ensure the validity of the tests is to check that all items have similar results across different groups of individuals. However, differential item functioning (DIF) occurs when the results of individuals with equal ability levels from different groups differ from each other on the same test item. Based on Item Response Theory and Classic Test Theory, there are some methods, with different advantages and limitations to identify items that show DIF. This study aims to compare the performances of five methods for detecting DIF. The efficacies of Mantel-Haenszel (MH), Logistic Regression (LR), Crossing simultaneous item bias test (CSIBTEST), Lord's chi-square (LORD), and Raju's area measure (RAJU) methods are examined considering conditions of the sample size, DIF ratio, and test length. In this study, to compare the detection methods, power and Type I error rates are evaluated using a simulation study with 100 replications conducted for each condition. Results show that LR and MH have the lowest Type I error and the highest power rate in detecting uniform DIF. In addition, CSIBTEST has a similar power rate to MH and LR. Under DIF conditions, sample size, DIF ratio, test length and their interactions affect Type I error and power rates.

Comparing Robust Linking and Regularized Estimation for Linking Two Groups in the 1PL and 2PL Models in the Presence of Sparse Uniform Differential Item Functioning

In the social sciences, the performance of two groups is frequently compared based on a cognitive test involving binary items. Item response models are often utilized for comparing the two groups. However, the presence of differential item functioning (DIF) can impact group comparisons. In order to avoid the biased estimation of groups, appropriate statistical methods for handling differential item functioning are required. This article compares the performance-regularized estimation and several robust linking approaches in three simulation studies that address the one-parameter logistic (1PL) and two-parameter logistic (2PL) models, respectively. It turned out that robust linking approaches are at least as effective as the regularized estimation approach in most of the conditions in the simulation studies.

Differential item functioning detection in fundamental physics test

The study's objective was to identify the differential item functioning (DIF) or bias using item response theory. This research is exploratory research with a quantitative approach. In this study, the data used is a fundamental physics test, specifically motion one and two dimensions, dynamics and work and energy concept, with 25 items in the two-tier multiple choices in form. Two hundred and fifty-four prospective physics teachers at West Java and Banten provinces, Indonesia, consisting of 103 males (reference group) and 151 females (focal group). The data analysis includes two stages: conformity of the logistic model parameters and item characteristics, followed by DIF detection. The model suitability test shows the appropriate parameter is 2PL. DIF detection has three methods: simple area indices, Raju and the Lord method. Research results show two things. First, the model fit test shows that the data best fit the 2-parameter or 2-PL logistic parameter model, and all items have good characteristics. Second, the most significant item containing uniform DIF was number 5, followed by nonuniform DIF items 9 and 25. These three items were biased because the chances of answering correctly for male and female students differed. Significantly, this means that this item favors one group. This study illustrates the importance of constructing instrument items so that all test takers, both male and female, have an equal chance to answer the question correctly. Keywords: Differential item functioning, physics test, simple area indices, Raju method, Lord method;