Kristin Jankowsky
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Kristin Jankowsky has a total of 516 co-authors. Here are the top co-authors by number of shared publications:
- Ulrich Schroeders
Department of Psychology, University of Kassel
8 shared publications - Gabriel Olaru
Differential Psychology and Psychological Assessment, Charlotte Fresenius Hochschule
3 shared publications - Ivan Ropovik
Faculty of Education, Charles University, Prague, Czech Republic
3 shared publications - Jamie Cummins
Institute of Psychology, University of Bern, Bern, Switzerland
3 shared publications - Matúš Adamkovič
Centre of Social and Psychological Sciences, Slovak Academy of Sciences 6 , Slovakia
3 shared publications Ulrich Schroeders
Psychological Assessment, University of Kassel, Kassel, Germany
3 shared publications
Latest Publications
- 10 September 2026
- Published by Springer Nature in Nature Human Behaviour
- Vol. 10 (10) , 1857-1859
How Registered Reports can improve trust in machine learning research
- 10 September 2026
- Published by Springer Nature in Nature Human Behaviour
- Vol. 10 (10) , 1857-1859
How Registered Reports can improve trust in machine learning research
- Open Access
- 30 April 2026
- Published by American Psychological Association (APA) in Psychological Methods
- Vol. 31 (5) , 1196-1217
Beyond the hype: A simulation study evaluating the predictive performance of machine learning models in psychology.
Although machine learning (ML) methods are gaining popularity in psychological research, the debate about their usefulness ranges from hype to disillusionment. The discrepancy between the hopes placed in ML methods and the empirical reality is often attributed to the quality of psychological data sets, which tend to be small and subject to imprecise measurement. In this simulation study, we examined the data requirements necessary for ML methods to perform well. We compared the performance of elastic net regressions with and without prespecified interactions, random forests, and gradient boosting machines for different data-generating processes (including interaction, stepwise, or piecewise linear effects) and under various conditions: (a) sample size, (b) number of irrelevant predictors, (c) predictor reliability, (d) effect size, and (e) nature of the data-generating process (i.e., linear vs. nonlinear effects). We investigated whether the models achieved the highest level of predictive performance attainable under the given simulated conditions. There were two main takeaways from our results: First, the maximum possible predictive performance was only achieved under optimal simulation conditions (N = 1,000, perfectly reliable predictors, predominantly linear effects, and an exceptionally large effect size of R² = .80), which are arguably rarely met in psychological research. Second, each ML model outperformed the others under certain conditions, but none was consistently superior or entirely robust to suboptimal data characteristics. We stress that data quality fundamentally limits predictive performance and discuss the interpretation of comparisons between flexible ML models and simpler (regularized linear) baselines in psychological research. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
- 30 April 2026
- Published by American Psychological Association (APA) in Psychological Methods
- Vol. 31 (5) , 1196-1217
Open AccessBeyond the hype: A simulation study evaluating the predictive performance of machine learning models in psychology.
Although machine learning (ML) methods are gaining popularity in psychological research, the debate about their usefulness ranges from hype to disillusionment. The discrepancy between the hopes placed in ML methods and the empirical reality is often attributed to the quality of psychological data sets, which tend to be small and subject to imprecise measurement. In this simulation study, we examined the data requirements necessary for ML methods to perform well. We compared the performance of elastic net regressions with and without prespecified interactions, random forests, and gradient boosting machines for different data-generating processes (including interaction, stepwise, or piecewise linear effects) and under various conditions: (a) sample size, (b) number of irrelevant predictors, (c) predictor reliability, (d) effect size, and (e) nature of the data-generating process (i.e., linear vs. nonlinear effects). We investigated whether the models achieved the highest level of predictive performance attainable under the given simulated conditions. There were two main takeaways from our results: First, the maximum possible predictive performance was only achieved under optimal simulation conditions (N = 1,000, perfectly reliable predictors, predominantly linear effects, and an exceptionally large effect size of R² = .80), which are arguably rarely met in psychological research. Second, each ML model outperformed the others under certain conditions, but none was consistently superior or entirely robust to suboptimal data characteristics. We stress that data quality fundamentally limits predictive performance and discuss the interpretation of comparisons between flexible ML models and simpler (regularized linear) baselines in psychological research. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
- Top Cited Papers
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 126-134
Investigating the reproducibility of the social and behavioural sciences
Published claims should be reproducible, yielding the same result when the same analysis is applied to the same data1,2. Here we assess reproducibility in a stratified random sample of 600 papers published from 2009 to 2018 in 62 journals spanning the social and behavioural sciences. The authors of 144 (24.0%, 95% confidence interval (CI) = 20.8-27.6%) papers made data available to assess reproducibility and, for 38 others, we obtained source data to reconstruct the dataset. We assessed 143 out of the 182 available datasets and found that 76.6 (53.6%, 95% CI = 45.8-60.7%) papers were rated as precisely reproducible and 105.0 (73.5%, 95% CI = 66.4-80.0%) were rated as at least approximately reproducible (within 15% of the original effects or within 0.05 of original P values) after inverse weighting each of the 551 claims by the number of claims per paper. We observed higher reproducibility for papers from political science and economics compared with other fields, for more recent papers compared with older papers and for papers from journals that require data sharing. Implementation of measures to verify that research is reproducible is needed to support trustworthiness in the complex enterprise of knowledge production3,4.
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 126-134
Top Cited PapersInvestigating the reproducibility of the social and behavioural sciences
Published claims should be reproducible, yielding the same result when the same analysis is applied to the same data1,2. Here we assess reproducibility in a stratified random sample of 600 papers published from 2009 to 2018 in 62 journals spanning the social and behavioural sciences. The authors of 144 (24.0%, 95% confidence interval (CI) = 20.8-27.6%) papers made data available to assess reproducibility and, for 38 others, we obtained source data to reconstruct the dataset. We assessed 143 out of the 182 available datasets and found that 76.6 (53.6%, 95% CI = 45.8-60.7%) papers were rated as precisely reproducible and 105.0 (73.5%, 95% CI = 66.4-80.0%) were rated as at least approximately reproducible (within 15% of the original effects or within 0.05 of original P values) after inverse weighting each of the 551 claims by the number of claims per paper. We observed higher reproducibility for papers from political science and economics compared with other fields, for more recent papers compared with older papers and for papers from journals that require data sharing. Implementation of measures to verify that research is reproducible is needed to support trustworthiness in the complex enterprise of knowledge production3,4.
- Top Cited Papers
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 143-150
Investigating the replicability of the social and behavioural sciences
Pursuing replicability — independent evidence for previous claims — is important for creating generalizable knowledge1,2. Here we attempted replications of 274 claims of positive results from 164 quantitative papers published from 2009 to 2018 in 54 journals in the social and behavioural sciences. Replications were high powered on average to detect the original effect size (median of 99.6%), used original materials when relevant and available, and were peer reviewed in advance through a standardized internal protocol. Replications showed statistically significant results in the original pattern for 151 of 274 claims (55.1% (95% confidence interval (CI) 49.2–60.9%)) and for 80.8 of 164 papers (49.3% (95% CI 43.8–54.7%)), weighed for replicating multiple claims per paper. We observed modest variation in replication rates across disciplines (42.5–63.1%), although some estimates had high uncertainty. The median Pearson’s r effect size was 0.25 (95% CI 0.21–0.27) for original studies and 0.10 (95% CI 0.09–0.13) for replication studies, an 82.4% (95% CI 67.8–88.2%) reduction in shared variance. Thirteen methods for evaluating replication success provided estimates ranging from 28.6% to 74.8% (median of 49.3%). Some decline in effect size and significance is expected based on power to detect original effects and regression to the mean because we replicated only positive results. We observe that challenges for replicability extend across social–behavioural sciences, illustrating the importance of identifying conditions that promote or inhibit replicability3,4.
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 143-150
Top Cited PapersInvestigating the replicability of the social and behavioural sciences
Pursuing replicability — independent evidence for previous claims — is important for creating generalizable knowledge1,2. Here we attempted replications of 274 claims of positive results from 164 quantitative papers published from 2009 to 2018 in 54 journals in the social and behavioural sciences. Replications were high powered on average to detect the original effect size (median of 99.6%), used original materials when relevant and available, and were peer reviewed in advance through a standardized internal protocol. Replications showed statistically significant results in the original pattern for 151 of 274 claims (55.1% (95% confidence interval (CI) 49.2–60.9%)) and for 80.8 of 164 papers (49.3% (95% CI 43.8–54.7%)), weighed for replicating multiple claims per paper. We observed modest variation in replication rates across disciplines (42.5–63.1%), although some estimates had high uncertainty. The median Pearson’s r effect size was 0.25 (95% CI 0.21–0.27) for original studies and 0.10 (95% CI 0.09–0.13) for replication studies, an 82.4% (95% CI 67.8–88.2%) reduction in shared variance. Thirteen methods for evaluating replication success provided estimates ranging from 28.6% to 74.8% (median of 49.3%). Some decline in effect size and significance is expected based on power to detect original effects and regression to the mean because we replicated only positive results. We observe that challenges for replicability extend across social–behavioural sciences, illustrating the importance of identifying conditions that promote or inhibit replicability3,4.
- Open Access
- 17 June 2025
- Published by SAGE Publications in International Journal of Behavioral Development
- Vol. 50 (1) , 126-139
Predicting juvenile delinquency and criminal behavior in adulthood using machine learning
By violating social norms, deviant behavior is an important issue that affects society as a whole and has serious consequences for its individuals. Different scientific disciplines have proposed theories of deviant behavior that often fall short of predicting actual behavior. In this registered report, we used data from the longitudinal National Study of Adolescent to Adult Health (Add Health) to examine the predictability of juvenile delinquency (Wave I) and adult criminal behavior (Wave V), distinguishing between drug, property, and violent offenses. Comparing the predictive accuracy of traditional regression models with different machine learning algorithms (elastic net regression and gradient boosting machines), we found the elastic net regressions with item-level data performed best. The prediction of juvenile delinquency was relatively accurate for drug offenses ( R ² = .57), violent offenses ( R ² = .44), and property offenses ( R ² = .39), while the performance declined significantly for adult delinquency, with R ² values ranging from .16 to .13. Key predictors of juvenile delinquency versus adult criminal behavior were clearly different from each other. Early risk factors for adult criminal behavior included prior juvenile delinquency, particularly drug-related offenses, sex, and school-related issues such as suspension or expulsion. We discuss the findings in the context of relevant theories on the causes and development of criminal behavior and explore potential approaches for prevention and early intervention, particularly within the framework of the “Central Eight.”
- 17 June 2025
- Published by SAGE Publications in International Journal of Behavioral Development
- Vol. 50 (1) , 126-139
Open AccessPredicting juvenile delinquency and criminal behavior in adulthood using machine learning
By violating social norms, deviant behavior is an important issue that affects society as a whole and has serious consequences for its individuals. Different scientific disciplines have proposed theories of deviant behavior that often fall short of predicting actual behavior. In this registered report, we used data from the longitudinal National Study of Adolescent to Adult Health (Add Health) to examine the predictability of juvenile delinquency (Wave I) and adult criminal behavior (Wave V), distinguishing between drug, property, and violent offenses. Comparing the predictive accuracy of traditional regression models with different machine learning algorithms (elastic net regression and gradient boosting machines), we found the elastic net regressions with item-level data performed best. The prediction of juvenile delinquency was relatively accurate for drug offenses ( R ² = .57), violent offenses ( R ² = .44), and property offenses ( R ² = .39), while the performance declined significantly for adult delinquency, with R ² values ranging from .16 to .13. Key predictors of juvenile delinquency versus adult criminal behavior were clearly different from each other. Early risk factors for adult criminal behavior included prior juvenile delinquency, particularly drug-related offenses, sex, and school-related issues such as suspension or expulsion. We discuss the findings in the context of relevant theories on the causes and development of criminal behavior and explore potential approaches for prevention and early intervention, particularly within the framework of the “Central Eight.”
- Open Access
- 8 October 2024
- Published by Hogrefe Publishing Group in European Journal of Psychological Assessment
- Vol. 42 (3) , 298-307
Is It Just a Game?
Building on the long common history of board games and intelligence research, we developed a new deductive reasoning test based on the popular game Mastermind. The research questions of this registered report were: (a) Is a psychometrically sound measurement of the ability to solve Mastermind items possible (i.e., a reliable, uni-dimensional measurement with a good coverage of difficulty)? (b) Is the ability to solve Mastermind items substantially related to other measures of cognitive ability (i.e., matrix test, knowledge test) and need for cognition? (c) Can item difficulty be predicted by the number of colors, positions, premises, and a newly proposed entropy-based index? Based on the results of a pilot study, we developed 30 items and administered them to 351 participants in the preregistered main study using a multiple matrix sampling design. The deductive Mastermind test proved to be (a) a reliable and efficient measure of reasoning across a wide ability range, and (b) showed expectation-consistent patterns for the convergent and divergent measures. (c) The entropy-based index allowed for the prediction of item difficulty to a considerable degree ( R2 = .47). We discuss the ideas of information theory, including entropy, as constructing principles for the rational test development of reasoning tests.
- 8 October 2024
- Published by Hogrefe Publishing Group in European Journal of Psychological Assessment
- Vol. 42 (3) , 298-307
Open AccessIs It Just a Game?
Building on the long common history of board games and intelligence research, we developed a new deductive reasoning test based on the popular game Mastermind. The research questions of this registered report were: (a) Is a psychometrically sound measurement of the ability to solve Mastermind items possible (i.e., a reliable, uni-dimensional measurement with a good coverage of difficulty)? (b) Is the ability to solve Mastermind items substantially related to other measures of cognitive ability (i.e., matrix test, knowledge test) and need for cognition? (c) Can item difficulty be predicted by the number of colors, positions, premises, and a newly proposed entropy-based index? Based on the results of a pilot study, we developed 30 items and administered them to 351 participants in the preregistered main study using a multiple matrix sampling design. The deductive Mastermind test proved to be (a) a reliable and efficient measure of reasoning across a wide ability range, and (b) showed expectation-consistent patterns for the convergent and divergent measures. (c) The entropy-based index allowed for the prediction of item difficulty to a considerable degree ( R2 = .47). We discuss the ideas of information theory, including entropy, as constructing principles for the rational test development of reasoning tests.
- Open Access
- 1 October 2024
- Published by Hogrefe Publishing Group in European Journal of Psychological Assessment
- Vol. 40 (5) , 376-395
Short-Scale Construction Using Meta-Analytic Ant Colony Optimization
The Need for Cognition Scale (NCS) is a self-report scale measuring individual differences in the tendency to engage in and enjoy thinking. The shortened version with 18 items (NCS-18; Cacioppo et al., 1984 ) has widely been administered in research on persuasion, critical thinking, and educational achievement. Whereas most studies advocated for essential uni-dimensionality, the question remains which psychometric model yields the best representation of the NCS-18. In the present study, we compared six competing measurement models for the NCS-18 with meta-analytic structural equation models using summary data of 87 samples ( N = 90,215). Results demonstrated that the negatively worded items introduced considerable measurement bias that was best accounted for with an acquiescence model. In a further analytical step, we showcased how the pooled correlation matrix can be used to compile short versions of the NCS-18 via Ant Colony Optimization. We examined model fit and reliability of short scales with varying item numbers (between 4 and 15) and a balanced ratio of positively and negatively worded items. We discuss the potentials and limits of the newly proposed method.
- 1 October 2024
- Published by Hogrefe Publishing Group in European Journal of Psychological Assessment
- Vol. 40 (5) , 376-395
Open AccessShort-Scale Construction Using Meta-Analytic Ant Colony Optimization
The Need for Cognition Scale (NCS) is a self-report scale measuring individual differences in the tendency to engage in and enjoy thinking. The shortened version with 18 items (NCS-18; Cacioppo et al., 1984 ) has widely been administered in research on persuasion, critical thinking, and educational achievement. Whereas most studies advocated for essential uni-dimensionality, the question remains which psychometric model yields the best representation of the NCS-18. In the present study, we compared six competing measurement models for the NCS-18 with meta-analytic structural equation models using summary data of 87 samples ( N = 90,215). Results demonstrated that the negatively worded items introduced considerable measurement bias that was best accounted for with an acquiescence model. In a further analytical step, we showcased how the pooled correlation matrix can be used to compile short versions of the NCS-18 via Ant Colony Optimization. We examined model fit and reliability of short scales with varying item numbers (between 4 and 15) and a balanced ratio of positively and negatively worded items. We discuss the potentials and limits of the newly proposed method.
- Open Access
- 18 December 2023
- Published by Wiley in British Journal of Clinical Psychology
- Vol. 63 (2) , 137-155
Predicting treatment response using machine learning: A registered report
Objective: Previous research on psychotherapy treatment response has mainly focused on outpatients or clinical trial data which may have low ecological validity regarding naturalistic inpatient samples. To reduce treatment failures by proactively screening for patients at risk of low treatment response, gain more knowledge about risk factors and to evaluate treatments, accurate insights about predictors of treatment response in naturalistic inpatient samples are needed. Methods: We compared the performance of different machine learning algorithms in predicting treatment response, operationalized as a substantial reduction in symptom severity as expressed in the Patient Health Questionnaire Anxiety and Depression Scale. To achieve this goal, we used different sets of variables—(a) demographics, (b) physical indicators, (c) psychological indicators and (d) treatment‐related variables—in a naturalistic inpatient sample ( N = 723) to specify their joint and unique contribution to treatment success. Results: There was a strong link between symptom severity at baseline and post‐treatment ( R 2 = .32). When using all available variables, both machine learning algorithms outperformed the linear regressions and led to an increment in predictive performance of R 2 = .12. Treatment‐related variables were the most predictive, followed psychological indicators. Physical indicators and demographics were negligible. Conclusions: Treatment response in naturalistic inpatient settings can be predicted to a considerable degree by using baseline indicators. Regularization via machine learning algorithms leads to higher predictive performances as opposed to including nonlinear and interaction effects. Heterogenous aspects of mental health have incremental predictive value and should be considered as prognostic markers when modelling treatment processes.
- 18 December 2023
- Published by Wiley in British Journal of Clinical Psychology
- Vol. 63 (2) , 137-155
Open AccessPredicting treatment response using machine learning: A registered report
Objective: Previous research on psychotherapy treatment response has mainly focused on outpatients or clinical trial data which may have low ecological validity regarding naturalistic inpatient samples. To reduce treatment failures by proactively screening for patients at risk of low treatment response, gain more knowledge about risk factors and to evaluate treatments, accurate insights about predictors of treatment response in naturalistic inpatient samples are needed. Methods: We compared the performance of different machine learning algorithms in predicting treatment response, operationalized as a substantial reduction in symptom severity as expressed in the Patient Health Questionnaire Anxiety and Depression Scale. To achieve this goal, we used different sets of variables—(a) demographics, (b) physical indicators, (c) psychological indicators and (d) treatment‐related variables—in a naturalistic inpatient sample ( N = 723) to specify their joint and unique contribution to treatment success. Results: There was a strong link between symptom severity at baseline and post‐treatment ( R 2 = .32). When using all available variables, both machine learning algorithms outperformed the linear regressions and led to an increment in predictive performance of R 2 = .12. Treatment‐related variables were the most predictive, followed psychological indicators. Physical indicators and demographics were negligible. Conclusions: Treatment response in naturalistic inpatient settings can be predicted to a considerable degree by using baseline indicators. Regularization via machine learning algorithms leads to higher predictive performances as opposed to including nonlinear and interaction effects. Heterogenous aspects of mental health have incremental predictive value and should be considered as prognostic markers when modelling treatment processes.
- Open Access
- 4 May 2023
- Published by SAGE Publications in European Journal of Personality
- Vol. 38 (2) , 306-325
Structural and Inter-individual Differentiation in Personality Traits Across the Adult Lifespan
The concept of differentiation describes increasing or decreasing similarities between inter-individual differences on psychological constructs, reflecting processes of specialization or adaptation. In this study, we studied age-differentiation in personality traits in (1) the trait domain and facet loadings, (2) the correlations between trait domains, and (3) trait domain, facet, and item (residual) variances. We used three large cross-sectional samples (Ns > 3000) covering 16–90 years of age with broad measures of the Big Five, Five-Factor, and HEXACO models. We examined age effects on the model parameters using local structural equation modeling. We found a high stability of the trait domain loadings, suggesting relatively stable trait domain compositions across age. Extraversion-Openness correlations increased across age for all three measures, whereas an increase in the Extraversion-Agreeableness and decrease in the absolute Neuroticism-Extraversion correlations only replicated across the five-dimensional models. Inter-individual differences in personality were similar across age in the trait domains and facets but differed substantially for item residuals. In summary, the structure and individual differences in broad personality traits is relatively stable across the adult lifespan, with most age-differences only affecting the item level.
- 4 May 2023
- Published by SAGE Publications in European Journal of Personality
- Vol. 38 (2) , 306-325
Open AccessStructural and Inter-individual Differentiation in Personality Traits Across the Adult Lifespan
The concept of differentiation describes increasing or decreasing similarities between inter-individual differences on psychological constructs, reflecting processes of specialization or adaptation. In this study, we studied age-differentiation in personality traits in (1) the trait domain and facet loadings, (2) the correlations between trait domains, and (3) trait domain, facet, and item (residual) variances. We used three large cross-sectional samples (Ns > 3000) covering 16–90 years of age with broad measures of the Big Five, Five-Factor, and HEXACO models. We examined age effects on the model parameters using local structural equation modeling. We found a high stability of the trait domain loadings, suggesting relatively stable trait domain compositions across age. Extraversion-Openness correlations increased across age for all three measures, whereas an increase in the Extraversion-Agreeableness and decrease in the absolute Neuroticism-Extraversion correlations only replicated across the five-dimensional models. Inter-individual differences in personality were similar across age in the trait domains and facets but differed substantially for item residuals. In summary, the structure and individual differences in broad personality traits is relatively stable across the adult lifespan, with most age-differences only affecting the item level.
- Open Access
- 24 April 2023
- Published by SAGE Publications in Assessment
- Vol. 31 (3) , 557-573
Predicting Lifetime Suicide Attempts in a Community Sample of Adolescents Using Machine Learning Algorithms
Suicide is a major global health concern and a prominent cause of death in adolescents. Previous research on suicide prediction has mainly focused on clinical or adult samples. To prevent suicides at an early stage, however, it is important to screen for risk factors in a community sample of adolescents. We compared the accuracy of logistic regressions, elastic net regressions, and gradient boosting machines in predicting suicide attempts by 17-year-olds in the Millennium Cohort Study ( N = 7,347), combining a large set of self- and other-reported variables from different categories. Both machine learning algorithms outperformed logistic regressions and achieved similar balanced accuracies (.76 when using data 3 years before the self-reported lifetime suicide attempts and .85 when using data from the same measurement wave). We identified essential variables that should be considered when screening for suicidal behavior. Finally, we discuss the usefulness of complex machine learning models in suicide prediction.
- 24 April 2023
- Published by SAGE Publications in Assessment
- Vol. 31 (3) , 557-573
Open AccessPredicting Lifetime Suicide Attempts in a Community Sample of Adolescents Using Machine Learning Algorithms
Suicide is a major global health concern and a prominent cause of death in adolescents. Previous research on suicide prediction has mainly focused on clinical or adult samples. To prevent suicides at an early stage, however, it is important to screen for risk factors in a community sample of adolescents. We compared the accuracy of logistic regressions, elastic net regressions, and gradient boosting machines in predicting suicide attempts by 17-year-olds in the Millennium Cohort Study ( N = 7,347), combining a large set of self- and other-reported variables from different categories. Both machine learning algorithms outperformed logistic regressions and achieved similar balanced accuracies (.76 when using data 3 years before the self-reported lifetime suicide attempts and .85 when using data from the same measurement wave). We identified essential variables that should be considered when screening for suicidal behavior. Finally, we discuss the usefulness of complex machine learning models in suicide prediction.