Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations
Methodological experts suggest that psychological and educational researchers should employ appropriate methods for data-driven model exploration, such as Bayesian Model Averaging and regularized regression, instead of conventional hypothesis-driven testing, if they want to explore the best predicti...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-07-01
|
| Series: | Stats |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2571-905X/7/3/44 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850259364392206336 |
|---|---|
| author | Hyemin Han |
| author_facet | Hyemin Han |
| author_sort | Hyemin Han |
| collection | DOAJ |
| description | Methodological experts suggest that psychological and educational researchers should employ appropriate methods for data-driven model exploration, such as Bayesian Model Averaging and regularized regression, instead of conventional hypothesis-driven testing, if they want to explore the best prediction model. I intend to discuss practical considerations regarding data-driven methods for end-user researchers without sufficient expertise in quantitative methods. I tested three data-driven methods, i.e., Bayesian Model Averaging, LASSO as a form of regularized regression, and stepwise regression, with datasets in psychology and education. I compared their performance in terms of cross-validity indicating robustness against overfitting across different conditions. I employed functionalities widely available via <i>R</i> with default settings to provide information relevant to end users without advanced statistical knowledge. The results demonstrated that LASSO showed the best performance and Bayesian Model Averaging outperformed stepwise regression when there were many candidate predictors to explore. Based on these findings, I discussed appropriately using the data-driven model exploration methods across different situations from laypeople’s perspectives. |
| format | Article |
| id | doaj-art-ba6274cebfed4f5fbe37c1a4cc61b986 |
| institution | OA Journals |
| issn | 2571-905X |
| language | English |
| publishDate | 2024-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Stats |
| spelling | doaj-art-ba6274cebfed4f5fbe37c1a4cc61b9862025-08-20T01:55:52ZengMDPI AGStats2571-905X2024-07-017373274410.3390/stats7030044Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical ConsiderationsHyemin Han0Educational Psychology Program, University of Alabama, Tuscaloosa, AL 35487, USAMethodological experts suggest that psychological and educational researchers should employ appropriate methods for data-driven model exploration, such as Bayesian Model Averaging and regularized regression, instead of conventional hypothesis-driven testing, if they want to explore the best prediction model. I intend to discuss practical considerations regarding data-driven methods for end-user researchers without sufficient expertise in quantitative methods. I tested three data-driven methods, i.e., Bayesian Model Averaging, LASSO as a form of regularized regression, and stepwise regression, with datasets in psychology and education. I compared their performance in terms of cross-validity indicating robustness against overfitting across different conditions. I employed functionalities widely available via <i>R</i> with default settings to provide information relevant to end users without advanced statistical knowledge. The results demonstrated that LASSO showed the best performance and Bayesian Model Averaging outperformed stepwise regression when there were many candidate predictors to explore. Based on these findings, I discussed appropriately using the data-driven model exploration methods across different situations from laypeople’s perspectives.https://www.mdpi.com/2571-905X/7/3/44data-driven analysismodel explorationvariable selectionBayesian Model Averagingregularized regressionLASSO |
| spellingShingle | Hyemin Han Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations Stats data-driven analysis model exploration variable selection Bayesian Model Averaging regularized regression LASSO |
| title | Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations |
| title_full | Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations |
| title_fullStr | Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations |
| title_full_unstemmed | Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations |
| title_short | Bayesian Model Averaging and Regularized Regression as Methods for Data-Driven Model Exploration, with Practical Considerations |
| title_sort | bayesian model averaging and regularized regression as methods for data driven model exploration with practical considerations |
| topic | data-driven analysis model exploration variable selection Bayesian Model Averaging regularized regression LASSO |
| url | https://www.mdpi.com/2571-905X/7/3/44 |
| work_keys_str_mv | AT hyeminhan bayesianmodelaveragingandregularizedregressionasmethodsfordatadrivenmodelexplorationwithpracticalconsiderations |