|
Size: 2022
Comment:
|
Size: 3566
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 3: | Line 3: |
| #acl HlpLabGroup:read,write,delete,revert,admin All:read | #acl HlpLabGroup:read,write,delete,revert,admin |
| Line 17: | Line 17: |
| * e.g. [http://idiom.ucsd.edu/~rlevy/teaching/fall2008/lign251/one_page_of_main_concepts.pdf roger's summary] |
|
| Line 19: | Line 21: |
| o collinearity o overfitting o overly influential cases o overdispersion? o model quality (e.g. residuals for linear models) o building a model: adding/removing variables (also: interactions) |
* collinearity * overfitting * overly influential cases * overdispersion? * model quality (e.g. residuals for linear models) * building a model: adding/removing variables (also: interactions) |
| Line 26: | Line 28: |
| o outlier handling o centering o removing collinearity (e.g. PCA, residualization) o stratification (using subsets of data) |
* outlier handling * centering * removing collinearity (e.g. PCA, residualization) * stratification (using subsets of data) |
| Line 31: | Line 33: |
| o testing significance (SE-based tests vs. model comparison) o interpration of model output, e.g. interpreation of coefficients o (also: coding of variables) o follow-up tests |
* testing significance (SE-based tests vs. model comparison) * interpration of model output, e.g. interpretation of coefficients * (also: coding of variables) * follow-up tests |
| Line 51: | Line 53: |
| * Plotting coefficient summaries * issues with scales and standardization * Examples (just some plots, along with function names): * [http://idiom.ucsd.edu/~rlevy/papers/doyle-levy-2008-bls.pdf p.8 for fixed effect summary, p.9 for random effect summary] * I have some functions, too. * Plotting individual effect shapes with confidence intervals: * back-transforming predictor to original scale (inverting: rcs, pol, centering, scaling, log, etc.) * choosing scale for outcome (inverting log-transformed scales or not; probabilities vs. log-odds for logit models) * Examples (just some plots, along with function names): * For ordinary regression models: {{{plot.Design()}}} * For mixed models: {{{plotLMER.fnc()}}}, {{{my.plot.glmer()}}}; see [http://hlplab.wordpress.com/2009/01/19/plotting-effects-for-glmer-familybimomial-models/ example] * Plotting model fit: * Calibration plot: * issues with calibration plots for logit models * Examples (just some plots, along with function names): * For ordinary models: {{{plot.calibration.Design()}}} * For mixed models: {{{my.plot.glmerfit()}}}; see [http://hlplab.wordpress.com/2009/01/19/visualizing-the-quality-of-an-glmerfamilybinomial-model/ example] * Visualization of predictors contribution to model (model comparison): {{{plot.anova.Design()}}} == What can you do to advance development of procedures and visualizations? == cite =) |
Workshop on common issues and standard in ordinary and multilevel regression modeling
March 25, 2009, UC Davis
Goal of this workshop
in progress
Common issues in regression modeling and some solutions
Maybe develop cheat sheet? with step-by-step guidelines of some things one should make sure to do when developing a model? does baayen or harell have something like this?
e.g. [http://idiom.ucsd.edu/~rlevy/teaching/fall2008/lign251/one_page_of_main_concepts.pdf roger's summary]
- common issues in regression modeling
- collinearity
- overfitting
- overly influential cases
- overdispersion?
- model quality (e.g. residuals for linear models)
- building a model: adding/removing variables (also: interactions)
- some solutions to these problems for common model types
- outlier handling
- centering
- removing collinearity (e.g. PCA, residualization)
- stratification (using subsets of data)
- interpreting the model, making sure the model answers the question of interest:
- testing significance (SE-based tests vs. model comparison)
- interpration of model output, e.g. interpretation of coefficients
- (also: coding of variables)
- follow-up tests
- common issues in regression modeling
Some suggestions on how to present model results
What do readers need to know?
What do reviewers need to know?
- How to talk about effect sizes?
- absolute coefficient size (related to range of predictor) -- talking about effect ranges.
- relative coefficient size (related to its standard error)
- model improvement, partial R-square, etc.
- accuracy?
- How to back-translated common transformations of outcomes and predictors?
Written description of the model
Visualization
- Plotting coefficient summaries
- issues with scales and standardization
- Examples (just some plots, along with function names):
[http://idiom.ucsd.edu/~rlevy/papers/doyle-levy-2008-bls.pdf p.8 for fixed effect summary, p.9 for random effect summary]
- I have some functions, too.
- Plotting individual effect shapes with confidence intervals:
- back-transforming predictor to original scale (inverting: rcs, pol, centering, scaling, log, etc.)
- choosing scale for outcome (inverting log-transformed scales or not; probabilities vs. log-odds for logit models)
- Examples (just some plots, along with function names):
For ordinary regression models: plot.Design()
For mixed models: plotLMER.fnc(), my.plot.glmer(); see [http://hlplab.wordpress.com/2009/01/19/plotting-effects-for-glmer-familybimomial-models/ example]
- Plotting model fit:
- Calibration plot:
- issues with calibration plots for logit models
- Examples (just some plots, along with function names):
For ordinary models: plot.calibration.Design()
For mixed models: my.plot.glmerfit(); see [http://hlplab.wordpress.com/2009/01/19/visualizing-the-quality-of-an-glmerfamilybinomial-model/ example]
Visualization of predictors contribution to model (model comparison): plot.anova.Design()
- Calibration plot:
What can you do to advance development of procedures and visualizations?
cite =)