The Memory Failures of Everyday Questionnaire (MFE) is one of the most widely-used instruments to assess memory failures in daily life. The original scale has nine response options, making it difficult to apply; we created a three-point scale (0-1-2) with response choices that make it easier to administer. We examined the two versions' equivalence in a sample of 193 participants between 19 and 64 years of age. The test-retest reliability and internal consistency of the version we propose were also computed in a sample of 113 people. Several indicators attest to the two forms' equivalence: the correlation between the items' means (r = .94; p < .001) and the order of the items' frequencies (r = .92; p < .001). However, the correlation between global scores on the two forms was not very high (r = .67; p < .001). The results indicate this new version has adequate reliability and internal consistency (rxx' = .83; p < .001; α = .83; p < .001) equivalent to those of the MFE 1-9. The MFE 0-2 provides a brief, simple evaluation, so we recommend it for use in clinical practice as well as research.

One approach to the study of everyday memory failures is to use multiple-item questionnaires. The Memory Failures of Everyday (MFE) test is one of the most frequently used in Spain. Our objective is to provide normative data from the MFE in a sample of healthy, Spanish, adult participants for use in clinical practice. The sample consists of 647 employees at a large company ranging in age from 19-64 years-old. Everyday memory failures were evaluated by means of the MFE with the following response format: 0-2 (0 = never or rarely; 1 = occasionally, sometimes; 2 = frequently, often). Results: Mean MFE = 15.25 (SD = 7.50), range 0-40. Correlation with age: .133 (p = .001); and with years of education: - .059 (n.s.). A constant increase in MFE was not observed across age groups (F = 4, 59;p = .003, η2 = .02), but differences were revealed between the 19-29 and 40-49 age groups; no differences were observed between the remaining age groups. Only slight differences between men and women occurred, the women's mean being slightly higher than the men's, but the confidence intervals overlapped (F = 5, 71; p = .017, η2 = .01). These results indicate that age, years of education, and sex had no significant effects. In light of the above, the sample was viewed as a whole when computing the percentiles reported here.