 Download Download Presentation Download Presentation - - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
##### Presentation Transcript

1. Model Adequacy Running a Real Regression Analysis Testing Assumptions, Checking for Outliers, and More

2. Hold up • Before the data was collected, did you bother to do a power analysis to estimate the sample size needed? • Same ingredients are necessary for determining N as before: alpha, effect size, desired power/beta • See an example in the commentary

3. Linearity • One should obviously check to see if the relation is a linear one • Check the regression of the DV on the composite • Tests can be done which examine curvilinear possibilities and whether those would be viable

4. Normal distribution of residuals • Our normality assumption applies to the residuals • One can simply save them and plot a density curve/histogram • Often a quantile-quantile plot is readily available, and here we hope to find most of our data along a 45 degree line

5. Homoscedasticity • We can check a plot of the residuals vs our predicted values to get a sense of the spread along the regression line • We prefer to see kind of a blob about the zero line (our mean), with no readily discernable pattern • This would mean that the residuals don’t get overly large for certain areas of the regression line relative to others

6. Collinearity • Multiple regression is capable of analyzing data with correlated predictor variables • However, problems can arise from situations in which two or more variables are highly intercorrelated • Perfect collinearity • Occurs if predictors are linear functions of each other (ex., age and year of birth), when the researcher creates dummy variables for all values of a categorical variable rather than leaving one out, and when there are fewer observations than variables • No unique regression solution • Less than perfect (the usual problem) • Inflates standard errors and makes assessment of the relative importance of the predictors unreliable • Also means that a small number of cases potentially can affect results strongly

7. Collinearity • Simple and Multi- Collinearity • When two or more variables are highly correlated • Can be detected by looking at the zero order correlations • Better is to regress each predictor on all other variables and look for large R2s* • Although our estimates of our coefficients are not necessarily biased, they become inefficient • Jump around a lot from sample to sample

8. Collinearity diagnostics • Tolerance* • Proportion of a predictors’ variance not accounted for by other variables • Looking for tolerance values that are small, close to zero as problematic • Means they are not contributing anything new to the model • tolerance = 1/VIF • VIF • Variance inflation factor • Looking for VIF values that are large • E.g. individual VIF greater than 10 should be inspected • VIF=1/tolerance • Other Indicators of Collinearity • Eigenvalues • Small values, close to zero • Condition index • Large values (15+)

9. Dealing with collinearity • Collinearity not necessarily a problem if the goal is to predict, not explain • Inefficiency of coefficients may not pose a real problem • Larger N might help reduce standard error of our coefficients • Combine variables to create a composite, Remove variable • Must be theoretically feasible • Centering the data (subtracting the mean) • Interpretation of coefficients will change as variables are now centered on zero • Recognize its presence and live with the consequences

10. Independence of residuals • We need to have our data points be independent of one another, as we have in other statistical analyses • As an example, subsets of various demographic categories could in theory result in a lack of independence • Include the variable in the model • The Durbin-Watson statistic is usually examined • 0 to 4, 2 good, deviations from it are to be examined • < 1 (indicates positive serial correlation), >3 (negative) • Better, use a statistical test for it • Like many tests regarding assumptions, we would prefer not to have a significant result

11. Regression Diagnostics • Of course all of the previous information would be relatively useless if we are not meeting our assumptions and/or have overly influential data points • In fact, you shouldn’t be really looking at the results unless you test assumptions and look for outliers, even though this requires running the analysis to begin with • Various tools are available for the detection of outliers • Classical methods • Standardized Residuals (ZRESID) • Studentized Residuals (SRESID) • Studentized Deleted Residuals (SDRESID) • Ways to think about outliers • Leverage • Discrepancy • Influence • Thinking ‘robustly’

12. Regression Diagnostics • Standardized Residuals (ZRESID) • Standardized errors in prediction • Mean 0, Sd = std. error of estimate • To standardize, divide each residual by its s.e.e. • At best an initial indicator (e.g. the +2 rule of thumb), but because the case itself determines what the variance would be, almost useless • Studentized Residuals (SRESID) • Same thing but studentized residual recognizes that the error associated with predicting values far from the mean of X is larger than the error associated with predicting values closer to the mean of X • standard error is multiplied by a value that will allow the result to take this into account • Studentized Deleted Residuals (SDRESID) • Studentized in which the standard error is calculated with the case in question removed from the others

13. Regression Diagnostics • Mahalanobis’ Distance • Mahalanobis distance is the distance of a case from the centroid of the remaining points (point where the means meet in n-dimensional space) • Cook’s Distance • Identifies an influential data point whether in terms of predictor or DV • A measure of how much the residuals of all cases would change if a particular case were excluded from the calculation of the regression coefficients. • With larger (relative) values, excluding a case would change the coefficients substantially. • DfBeta • Change in the regression coefficient that results from the exclusion of a particular case • Note that you get DfBetas for each coefficient associated with the predictors

14. Regression Diagnostics • Leverage assesses outliers among the predictors • Mahalanobis distance • Relatively high Mahalanobis suggests an outlier on one or more variables • Discrepancy • Measures the extent to which a case is in line with others • Influence • A product of leverage and discrepancy • How much would the coefficients change if the case were deleted? • Cook’s distance, dfBetas

15. Outliers • Influence plots • With a couple measures of ‘outlierness’ we can construct a scatterplot to note especially problematic cases • After fitting a regression model in R-commander, i.e. running the analysis, this graph is available via point and click • Here we have what is actually a 3-d plot, with 2 outlier measures on the x and y axes (studentized residuals and ‘hat’ values, a measure of leverage) and a third in terms of the size of the circle (Cook’s distance) • For this example, case 35 appears to be a problem

16. Summary: Outliers • No matter the analysis, some cases will be the ‘most extreme’. However, none may really qualify as being overly influential. • Whatever you do, always run some diagnostic analysis and do not ignore influential cases • It should be clear to interested readers whatever has been done to deal with outliers • As noted before, the best approach to dealing with outliers when they do occur is to run a robust regression with capable software

17. Suppressor variables • There are a couple of ways in which suppression can occur or be talked of, but the gist is that this masks the impact the predictor would have on the dependent if the third variable did not exist • In general suppression occurs when i falls outside the range of 0 ryi • Suppression in MR can entail some different relationships among predictors • For example one suppressor relationship would be where two variables, X1 and X2, are positively related to Y, but when the equation comes out we get • Y-hat = b1X1 – b2X2 + a • Three kinds to be discussed • Classical • Net • Cooperative

18. Suppression: Technical side • When dealing with standardized regression coefficients, note that

19. b.  Partial independence: R2Y.12 = 0 but  r12 0, a.  Complete independence: R2Y.12 = 0 d.  Partial independence again, both rY1 and rY2 ≠ 0, but r12 = 0 Suppression • Consider the following relationships

20. Suppression • e. Normal situation, redundancy: no simple correlation = 0 • Each semi-partial correlation, and the corresponding beta, will be less than the simple correlation between Xi and Y. This is because the variables share variance and influence • f.  Classical suppression:  rY2 = 0

21. Suppression • Recall from previously  • If ry2 = 0, then  • With increasingly shared variance between X1 and X2 we will have an inflated beta coefficient for X1 • X2 is suppressing the error variance in X1 • In other words, even though X2 is not correlated with Y, having it in the equation raises the R2 from what it would have been with just X1.

22. Suppression • Other suppression situations • Net • All rs positive • 2 ends up with a sign opposite that of its simple correlation with Y • It is always the X which has the smaller ryi which ends up with a  of opposite sign •  falls outside of the range 0 ryi, which is always true with any sort of suppression • Cooperative • Predictors negatively correlated with one another, both positive with DV • Or positively with one another and negatively with Y • Example of Cooperative: • Correlation between social aggressiveness (X1) and sales success (Y) = .29 • Correlation between record keeping (X2) and sales success (Y) = .24 • r12 = -.30 • Regression coefficients for predictors = .398 and .359 respectively

23. Suppression • Gist: weird stuff can happen in MR, so take note of the relationship of the predictors and how it may affect your overall interpretation • Compare the simple correlations of each predictor with the DV and compare to their respective beta coefficients* • If coefficient noticeably larger than simple correlation (absolute value) or of opposite sign one should suspect possible suppression

24. Model Validation • Overfitting • Validation • Bootstrapping

25. Overfitting • External validity • In some cases, some of the variation the parameters chosen are explaining is variation that is idiosyncratic to the sample • We would not see this variability in the population • So the fit of the model is good, but it doesn’t generalize as well as one would think • Capitalization on chance

26. Overfitting • Example from Lattin, Carroll, Green • Randomly generated 30 variables to predict an outcome variable • Using a best subsets approach, 3 variables were found that produce an R2 of .33 or 33% variance accounted for • As one can see, even random data has the capability of appearing to be a decent fit

27. Validation • One way to deal with such a problem is with a simple random split • With large datasets one can randomly split the sample into two sets • Calibration sample: used to estimate the coefficients • Holdout sample: used to validate the model • Some suggest a 2:1 or 4:1 split • This would require typically large samples for the holdout sample to be viable • Using the coefficients from the calibration set one can create predicted values for the holdout set • The squared correlation between the predicted values and observed values can then be compared to the R2 of the calibration set • In previous example of randomly generated data the R2 for the holdout set was 0

28. Other approaches • Jackknife Validation • Create estimates with a particular case removed • Use the coefficients obtained from analysis of the n-1 remaining cases to create a predicted value for the case removed • Do for all cases, and then compare the jackknifed R2 to the original • Subsets approach • Create several samples of the data of roughly equal size • Use the holdout approach with one sample, and obtain estimates from the others • Do this for each sample, obtain average estimates

29. Bootstrap • With relatively smaller samples*, cross-validation may not be as feasible • One may instead resample (with replacement) from the original data to obtain estimates for the coefficients • Use what is available to create a sampling distribution of for the values of interest

30. Summary • There is a lot to consider when performing multiple regression analysis • Actually running the analysis is just the first step, and if that’s all we are doing, we haven’t done much • Inferences are likely incomplete at best, innaccuarate at worst • A lot of work will be necessary to make sure that the conclusions drawn will be worthwhile • And that’s ok, you can do it!

31. Summary of how one could do regression • Idea pops into your head • Have some loose hypotheses about correlations among some variables • Collect some data • Run the regression analysis • Use R2 and standard-fare metrics of variable importance • Rely on statistical significance when you don’t have any real effects to talk about • This would be a bad way to do regression.

32. Summary of how to do a real regression analysis • 1. Have an idea • 2. Propose a theoretical (possibly causal) model in which you have thought about other viable models (including how predictors might predict one another, moderating and mediating possibilities* etc.) • 3. Collect appropriate and enough data (based on a power analysis) • Must have reliable measures • 4. Spend time with initial examination of data including obtaining a healthy understanding of the variables descriptively, missing values analysis if necessary, inspection of correlations etc. • 5. Run the analysis. Might as well ignore for now. • 6. With the model in place, test assumptions, look for collinearity, identify outliers. Take appropriate steps necessary to deal with any issues including bootstrapped regression or robust regression • 7. Rerun the analysis. • 8. Validate the model. Note any bias. • 9. Interpret results. Focus on bias corrected estimates of R2, interval estimates of coefficients, interpretable measures of variable importance (test for differences on them)