1 / 19

Class 8

Class 8. Indexes and Scales. Class Outline. Indexes versus Scales Guttman Scale Index Construction CES-D (measure of depression) Segregation index. Index and Scale. Similarities: Both are composite measurements, i.e., measurements based on multiple items.

benard
Download Presentation

Class 8

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Class 8 Indexes and Scales

  2. Class Outline • Indexes versus Scales • Guttman Scale • Index Construction • CES-D (measure of depression) • Segregation index

  3. Index and Scale Similarities: • Both are composite measurements, i.e., measurements based on multiple items. • Both are ordinal measures, which rank order units of analysis in terms of specific variables. • Both are often used as interval measures in practice. That is, we treat them as continuous variables.

  4. Index and Scale Difference: • Index: accumulate scores assigned to individual attributes. • Unweighted: Index = score1 + score2 + score3 + … • Weighted: Index= w1*score1 + w2*score2 + w3*score3 + … • Scale: assign scores to patterns of responses.

  5. Index and Scale Why do we need them? • Inadequacy of single indicators Single indicators are often insufficient for abstract or latent (unobservable) variables such as alienation, religiosity, prejudice, well-being, quality of life, ability, etc. • Data reduction Index and scale construction is an efficient way to summarize data when we have many variables to work with but want a clear, sharp result. (e.g., GPA)

  6. Guttman Scale Response Patterns The difficulty levels of the five math problems follow a clear pattern.

  7. Bogardus Social Distance Measure Response Patterns

  8. Constructing an Index • Select items for a composite index. • Examine empirical relationships. • Assign scores for responses. • Handle missing data.

  9. An Example: Clinical Measure of Depression--the CES-D Scale • On how many days during the past week did you feel you could not shake off the blues even with help from your family and friends? [0, 1, 2, 3, 4, 5, 6, 7] • On how many days during the past week did you feel bothered by things that don't usually bother you? [0, 1, 2, 3, 4, 5, 6, 7] • On how many days during the past week did you think your life had been a failure? [0, 1, 2, 3, 4, 5, 6, 7] • On how many days during the past week did you feel happy? [0, 1, 2, 3, 4, 5, 6, 7] • On how many days during the past week did you feel that people were unfriendly? [0, 1, 2, 3, 4, 5, 6, 7] • …

  10. Examine Items . summarize Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- id | 6196 446073.6 183696.2 101003 771015 ces_d | 5318 16.06168 15.16228 0 126 numans | 5367 19.74511 1.88859 0 20 mu003rer | 5281 .3686802 .9919353 0 7 mu004rer | 5268 .5442293 1.007349 0 7 mu005rer | 5280 .2681818 .9086162 0 7 mu006rer | 5265 5.451472 1.688951 0 7 mu007rer | 5268 .6753986 1.153571 0 7 mu008rer | 5273 .7344965 1.432888 0 7 mu009rer | 5287 5.812559 1.743115 0 7 mu010rer | 5275 .1672038 .6882907 0 7 mu011rer | 5280 .3909091 .9510827 0 7 mu012rer | 5281 .8763492 1.340246 0 7 mu013rer | 5279 .6541012 1.263232 0 7 mu014rer | 5277 .8609058 1.300107 0 7 mu015rer | 5286 .2415815 .845546 0 7 mu016rer | 5260 5.615399 2.289262 0 7 mu017rer | 5279 .9414662 1.532823 0 7 mu018rer | 5281 5.207726 2.239078 0 7 mu019rer | 5261 .6565292 1.325413 0 7 mu020rer | 5286 1.393681 1.819465 0 7 mu021rer | 5278 .6661614 1.124079 0 7 mu022rer | 5277 .8487777 1.292901 0 7

  11. Histograms of Items 1-4: Data Source: The Wisconsin Longitudinal Study

  12. Checking Correlations Among Items

  13. Histogram of Total CES-D Score

  14. Index as a Weighted Mean • If we wish to give certain items more importance than other items, we use weights: • Index = S wi * scorei / S wi, • A typical weighted index is GPA. • Two assumptions are involved: • Interval scale. (A - B = B – C) • Equal importance for each credit hour. • GPA = S(credit hoursi*gradesi) / S credit hoursi

  15. Assign Scores for Responses • Decide the desirable range of the index scores. • Some items may need be “reverse-coded” before being added up. • Decide whether to give each item in the index equal weight or different weights.

  16. Ways to Handle Missing Data in Index Construction • Exclude cases with missing data from the construction of the index and the analysis. (We lose observations.) • Treat missing data as one of the available responses (usually as “no” or “neutral”). • Assign the variable average to missing cases. • Analyze missing data to interpret the meaning.

  17. Segregation Index Percent white population in the neighborhood Cumulative white population in the first X neighborhood Assuming that there are only 5 neighborhoods in each city and all neighborhoods have the same number of residents.

  18. Segregation Index • Index of dissimilarity is calculated as D= 0.5 S |Pig/Pg-Pih/Ph| • Pig is the population of group g in census tract i • Pih is the population of group h in census tract i • Pg is the total population of group g and • Ph is the total population of group h • Interpretation of D: the (minimum) proportion of group g or h that needs to be moved to different neighborhoods in order for the city to be evenly distributed.

  19. Segregation Index

More Related