This study tested the factorial and measurement invariance of the Mini Asthma Quality of Life Questionnaire (MiniAQLQ) in a sample of Latino and African–American asthmatic patients.
We used confirmatory factor analyses (CFA) to examine data from adult outpatients diagnosed with asthma (n = 265). We proceeded by examining the original factor structure of the MiniAQLQ in the whole sample. We then determined whether the factor structure supported by the data was invariant across African–American and Latino patients by testing a series of nested models.
Results revealed that a three-factor structure model that combined the emotional function and symptom domains had the best fit to the whole sample (χ 2(91) = 99.83, P = 0.08). Analysis using a MIMIC model among Latino participants indicated that survey language did not influence item functioning (χ 2(93) = 103.29, P = 0.22). Multi-group CFA showed that both the structure and the meaning of the items were invariant across Africa-American and Latino participants. Scalar and strict measurement invariance was supported for 11 out of 15 items.
These data suggest that structure of the MiniAQLQ needs to be reassessed. Additional research needs to examine whether elimination of current items or inclusion of new items can ensure that constructs are measured without bias across different ethnic groups.