Measuring Older Adults' Abuse: Evaluation of Formative Indicators to Promote Brevity

Abstract

Some literature on elder abuse recommends, and practitioners claim, that there should be better assessment and screening tools. In order to improve the accuracy of measurement instruments, the purpose of this article is threefold: (a) describing the construction of an instrument with formative indicators and the survey design about the sensitive topic of elder abuse, (b) development of an analytic strategy to improve the precision of the measures by (c) evaluating the measurement instrument through quality criteria against outcomes of the instrument. We randomly selected 2,880 home-dwelling older women aged 60 and above from five European Union countries who participated in a survey on elder abuse. Prevalence data on abuse against older women was gathered using a postal (BE, FI, PT), face-to-face (BE, LT), and telephone survey (AT) but using an identical instrument. A table with outcome measures was calculated to evaluate the formative indicators of the measurement instrument, and a decision strategy for item reduction was developed. The results suggest that 12 (35%) of the original 34-indicators instrument can be omitted. The adapted version can provide the same elder abuse prevalence rates (reliability) with the same negative associations in terms of life quality (validity). The results indicate in an applied way how an elder abuse instrument can be evaluated and further developed using formative measures.

Acknowledgments

There are no conflicts of interest and the views expressed here are those of the researchers and not the European Union. Special thanks are given to the older women in all five countries who participated in the study, for sharing their views and reporting their experiences.

Notes

^a no target person in household.

^b e.g., wrong/unknown address/telephone number, responded, moved/deceased, nonliving houses (e.g., shops, enterprises), impossible to reach the house, sampling point not found.

^c maximum three visits or calls were taken.

^d refusal of the target person (e.g., questionnaire not returned or returned blank) or of a third person.

^e e.g., respondent did not take the telephone call or opened the door.

^f e.g., due to health problems, foreign language.

Note. Bolded numbers are indicating grounds in the decision making process.

^a Indicators deleted in the shortened instrument.

*p < .01.