BDI-II - Administration matters

What is it about?

Self-reports: questionnaires, inventories and scales are routinely used to assess people in research or clinical practice as it is time-saving for the clinicians or researchers. The results are used in diagnostic process, or for study sample description or an inclusion criterion. The methods are usually developed for certain group (children, adults, seniors; clinical population, etc.), and certain administration (self-report, interview) but when they become popular their use is stretched also to other groups, and original interview-based methods are sometimes used as self-reports. Then, it can happen that the addressed population may not fully understand what the questions/items are about, as the wording was aimed at specific situations, which differed from theirs. E.g. methods developed for and with clinical populations, people having specific health issues (depression, stroke, etc.), often use wording implicating the state before the condition/impairment/disease/injury and now. This is irrelevant for healthy control subjects and may be confusing. We have made a similar experience when administering Beck Depression Inventory (BDI-II) developed originally for adult depressed patients to healthy older persons in a study on cognition. To assess study participants mental health, especially level of depressive symptomatology, is often done so to document that the sample is indeed healthy (non-depressed). Their questions during the assessment led us to realize that the wording of items including "as usual" etc. was ambiguous, and that even we were not sure about the right answer when they inquired what we want to know - whether they are supposed to compare present state with their youth, productive mid-life adulthood or just the short period of time mentioned in the instructions of BDI-II. Thus, we designed a short study to check the extent of the unreliability of the method used in this situation. First, we asked our study participants, mentally healthy community-dwelling older people, to fill-in BDI-II as a self-report (which is now the usual way of administration). They were instructed to read the standard instructions (saying they should consider last two weeks) before they start working on it. After a short break we went through it with them again, this time as a structured interview, which gave them opportunity to discuss the items. In interview, we stressed the instruction to limit the time of consideration only to the last two weeks. The total scores then dropped (88%). For example, 58% older persons changed scores in the item on loss of interest in sex. In the self-report, they gave higher score (i.e. worse) as they compared their interest in sex now and when they were in their prime youth. In the interview, it also became apparent that their sex drive is not as urgent as it was when their were adolescent or young, but they are still interested in sex and intimacy and quite happy with it (i.e. "not depressed" for that). Other items, especially the somatic ones, showed similar results.

Featured Image

Why is it important?

It is important to check the wording of items in questionnaires/scales and critically assess whether the person or the group I intend to administer it to, will be able to understand it, whether it is relevant for them, and whether I am actually able to answer their questions if they are in doubt - do I (a clinician or researcher) understand 100% the method I am about to administer? Automatic adoption of methods without such a critical consideration may give unreliable results. Authors of reports and articles should always mention the way of administration of the methods they used, because different administrations may give very different restults. It will also be good for replication studies to make this very clear.