Table 1.
Reliability Indicator Affected? |
||||||
---|---|---|---|---|---|---|
Absolute | Differential | Validity Affected? | ||||
Source | Alpha | Retest | Alpha | Retest | Absolute | Differential |
Item Irrelevance | Yes | No | Yes | No | Yes | Yes |
Item Heterogeneity | Yes | No | Yes | No | ? | ? |
State Variation | No | Yes | No | Yes | Yes | Yes |
Respondents' Error | Yes | Yes | No | No | Yes | No |
Item Ambiguity | Yes | Yes | Yes | Yes | Yes | Yes |
Sample Variance | Yes | Yes | Yes | Yes | Yes | Yes |
Note. This table summarizes conceptual arguments offered in the text, not empirical findings. Absolute = magnitude for a given test compared to a fixed standard (e.g., .70); Differential = magnitude compared to that of other tests administered to the same sample. Item irrelevance = inappropriateness of item content. Item heterogeneity = statistical independence of trait indicators. State variation = effects of time-of-administration. Respondents' error = item misunderstanding or carelessness. Item ambiguity = lack of clarity or intelligibility of items; also includes effects due to the interaction of respondent and item characteristics. Sample variability = range of variation of traits in the sample. ? = Effect unknown.