Gender differences on general knowledge tests: Are they due to Differential Item Functioning?

Steinmayr, Ricarda; Bergold, Sebastian; Margraf-Stiksrud, Jutta; Freund, Philipp Alexander

doi:10.1016/j.intell.2015.04.001

Cited by 17 publications

(13 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to compare our sample with the norm samples of the instruments, we also calculated the T values ( M = 50, SD = 10) and the standard deviations of the T values (for the personality factors, the norm samples for 16- to 20-year old men and women, respectively, from 2007 [56] were used, because there is no standardization for the German NEO-FFI edition from 1993; the instrument itself remained unchanged from 1993 to 2007). From the T values, it can be seen that the sample was slightly above average, and showed some variance restriction, in intelligence, which is typical for students from Gymnasiums [57,58,59]. Regarding personality, all the factors but O matched the mean of the norming samples quite well and did not show any greater restriction in variance.…”

Section: Studymentioning

confidence: 89%

Personality and Intelligence Interact in the Prediction of Academic Achievement

Bergold

Steinmayr

2018

J. Intell.

Self Cite

View full text Add to dashboard Cite

Personality predicts academic achievement above and beyond intelligence. However, studies investigating the possible interaction effects between personality and intelligence when predicting academic achievement are scarce, as is the separate investigation of broad personality factors versus narrow personality facets in this context. Two studies with 11th grade students (Study 1: N = 421; Study 2: N = 243) were conducted to close this research gap. The students completed the Intelligence-Structure-Test 2000 R measuring general reasoning ability, and a well-established personality inventory based on the Five Factor Model. Academic achievement was operationalized via Grade Point Average. Using hierarchical regression and moderation analyses, Study 1 revealed that Conscientiousness interacted with intelligence when predicting academic achievement: there was a stronger association between intelligence and academic achievement when students scored higher on the Conscientiousness scale. Study 2 confirmed the findings from Study 1 and also found a moderation effect of Neuroticism (stronger association between intelligence and academic achievement with lower values on the Neuroticism scale). Analyses at the facet level revealed much more differentiated results than did analyses at the domain level, suggesting that investigating personality facets should be preferred over investigating personality domains when predicting academic achievement.

show abstract

Section: Studymentioning

confidence: 89%

Personality and Intelligence Interact in the Prediction of Academic Achievement

Bergold

Steinmayr

2018

J. Intell.

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, as with measurement invariance and differential item functioning research, sometimes it is unclear why a violation occurs. This can then be addressed in follow-up research (e.g., explaining the invariance using covariates, see Steinmayr, Bergold, Margraf-Stiksrud, & Freund, 2015).…”

Section: Exploratory and Confirmatory Usementioning

confidence: 99%

Hidden Markov Item Response Theory Models for Responses and Response Times

Molenaar

Oberski

Vermunt

et al. 2016

Multivariate Behavioral Research

View full text Add to dashboard Cite

Current approaches to model responses and response times to psychometric tests solely focus on between-subject differences in speed and ability. Within subjects, speed and ability are assumed to be constants. Violations of this assumption are generally absorbed in the residual of the model. As a result, within-subject departures from the between-subject speed and ability level remain undetected. These departures may be of interest to the researcher as they reflect differences in the response processes adopted on the items of a test. In this article, we propose a dynamic approach for responses and response times based on hidden Markov modeling to account for within-subject differences in responses and response times. A simulation study is conducted to demonstrate acceptable parameter recovery and acceptable performance of various fit indices in distinguishing between different models. In addition, both a confirmatory and an exploratory application are presented to demonstrate the practical value of the modeling approach.

show abstract

“…Specifically, while CTT showed a gender gap in 81% of sample countries and territories, IRT shows a gender gap in only 54% (21 out of 39) of the sample. These results are consistent with other knowledge testing assessments which also suffer from DIF for example the gender gap in political knowledge (Lizotte and Sidman, 2009) and high school educational testing (Steinmayr et al ., 2015).…”

Section: Resultsmentioning

confidence: 99%

“…In a gender‐fair test, the relation between the ability being measured and test behavior should not be influenced by gender (Steinmayr et al ., 2015). Yet it is widely observed that women are more likely than men to choose the “do not know” response in a multiple choice test (Agnew et al ., 2008; Lusardi et al ., 2010; Bucher‐Koenen et al ., 2017; Chen and Garand, 2018).…”

Section: Literature Reviewmentioning

confidence: 99%

“…IRT methods have been employed for decades in a variety of testing domains from education (Steinmayr et al ., 2015) and political knowledge (Lizotte and Sidman, 2009) to health and psychology (Fleishman and Lawrence 2003; Fleishman et al, 2002). For the various reasons it is a dependable method in these other domains, I argue that IRT is more appropriate in the financial knowledge measurement context than CTT methods.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Give mind to the gap: Measuring gender differences in financial knowledge

Ooi

2020

Journal of Consumer Affairs

View full text Add to dashboard Cite

Prior research consistently finds a gender gap in financial knowledge where males appear to outperform females. Despite the wealth of studies attempting to explain this gap, none have considered whether the gender gap may be a product of measurement method. This study reexamines the gender gap with item response theory (IRT) which can account for guessing behavior and differential item functioning. Survey data on 184,869 individuals from 39 countries and territories is analyzed. Results show that when IRT is employed, a gender gap exists in only 54% of the sample. In contrast, when a conventional measurement approach is used, there is a gender gap in financial knowledge in 81% of the sample. These results reveal that prior measurements may underestimate women's financial knowledge and inflate the gender gap.

show abstract

Gender differences on general knowledge tests: Are they due to Differential Item Functioning?

Cited by 17 publications

References 36 publications

Personality and Intelligence Interact in the Prediction of Academic Achievement

Personality and Intelligence Interact in the Prediction of Academic Achievement

Hidden Markov Item Response Theory Models for Responses and Response Times

Give mind to the gap: Measuring gender differences in financial knowledge

Contact Info

Product

Resources

About