Reliability课件.PPT

Coefficient Alpha versus Split-half reliability Estimates… Split-Half Reliability – Divide the items on the assessment into 2 halves and then correlate the two halves. Problem: Estimates fluctuate depending on what items get split into which halves. Alpha is the average of all possible split-half reliabilities. 精品文档 Sample Matrix Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 John 4 3 5 5 3 2 Paul 4 5 5 3 4 4 Ringo 2 2 2 1 2 3 George 4 4 3 2 5 4 精品文档 Real Results 10 Item Measure of Self-Esteem for 451 women. Correlate the average of the odd number items with the average of the even number items: r = .79 Correlate average first five items with the average of the last five items: r = .67 Average Inter-Item r = .46 Standardized Alpha = .89 精品文档 Caveats about Coefficient Alpha …. Recall – what goes into the Alpha calculation: Number of items Average inter-item correlation There are at least two things to think about when considering Coefficient Alpha… Length of the Assessment Dimensionality 精品文档 Pay Attention to the Length of the Assessment. 精品文档 Constant average inter-item correlation (e.g. .420) but increase the number of items…. Items Standardized Alpha 5 .78 6 .81 7 .84 8 .85 9 .87 10 .88 100 .99 精品文档 Now let’s use something like an average inter-item correlation of .15 Items Standardized Alpha 5 .47 10 .64 15 .73 20 .78 25 .82 30 .84 100 .95 精品文档 With enough items it is possible to achieve very high alpha coefficients… 精品文档 Dimensionality of the Measure 精品文档 Let’s Get To The Same Average Inter-Item Correlation in Two Ways Example from Schmitt (1996) 精品文档 Item Pool #1 (Average inter-item r = .5) Item 1. 2. 3. 4. 5. 6. 1. 1.0 2. .8 1.0 3. .8 .8 1.0 4. .3 .3 .3 1.0 5. .3 .3 .3 .8 1.0 6. .3 .3 .3 .8 .8 1.0 精品文档 Item Pool #2 (Average inter-item r = .5) Item 1. 2. 3. 4. 5. 6. 1. 1.0 2. .5 1.0 3. .5 .5 1.0 4. .5 .5 .5 1.0 5. .5 .5 .5 .5 1.0 6. .5 .5 .5 .5 .5 1.0 精品文档 Let’s Calculate Alphas… Item Pool #1: Average Inter-item r = .50, number of items = 6. Item Pool #2: A