Year: 2005
Type: letter
Abstract: Examining the reasons that former heavy drinkers give for drinking less is a compelling research agenda [1]. There is hope that such exploration will provide a window into the motivations for change that are associated most strongly with resolving problem drinking. Also interesting is the extent to which the identified reasons for change are consistent or inconsistent across different research reports. To me, this variation implies that it is essential to interpret the results of these studies within the context of the particular research methods employed in each project. Such consideration should help to shed light on why different reasons are identified when different research methods are utilized, and highlight those reasons for change that are consistent no matter the research methods employed. A framework within which to consider reasons for change might contain the following dimensions: (a) retrospective or prospective design; (b) representative versus convenience sample; (c) reasons provided by researcher or self-generated by respondent; (d) severity of current or prior alcohol problems in respondent population; and (e) interpretation of reasons by the researcher. Are respondents asked about their reasons for change before they make a change or after the change is made [2]? Are reasons for change related to success at sustained resolutions? The difficulty with purely retrospective designs is that they may be more prone to attributional biases in which respondents attribute success to themselves and failures to situational factors [3]. The Matzger et al. study [1] is a combination of both a retrospective and a prospective design. Respondents were asked their reasons for change in the year after they had started drinking a great deal less. These reasons were then related prospectively to sustained success at remission. Did the study use a method that ensured a representative sample of the population of former or current heavy drinkers, whether the target population was a treatment service (or region of treatment services) or a sample from the community at large? Alternatively, did the researcher rely on the respondents self-selecting themselves in response to a newspaper advertisement or some other type of convenience sampling method? The issue to be aware of with convenience samples is that it may result in a group of respondents who have a story they want to tell about overcoming their problem drinking. Representative samples, in contrast, are more likely to contain former heavy drinkers who reduced their drinking but had no particular event or story to tell about it (the maturing-out hypothesis [4, 5]). The weakness that often occurs with representative designs, however, is that they usually do not allow for in-depth interviews regarding reasons for change. The Matzger et al. study [1] used a sophisticated sampling method to obtain a representative sample of respondents both from treatment agencies and from the community at large. Further, they employed face-to-face interviews that allowed for more in-depth explorations of factors of interest in their study. As Matzger et al. [1] note, providing respondents with a list of reasons for change has the strength of reducing the likelihood that respondents will reconstruct or re-interpret their reasons for change. The potential weakness of providing a list of reasons (as was carried out in the present study) is that it has the danger of limiting the potential reasons for change that respondents might provide. Is the list of reasons comprehensive? What was their source (extant literature, focus groups, etc.)? What happens for those respondents who really do not have any particular reason for changing their drinking? Do they endorse some provided reasons for change when in fact they just drifted out of heavy drinking? Rumpf and colleagues [6] have pointed out that media-solicited natural history studies result in samples of respondents who, on average, have more severe problems than those samples recruited using population sampling techniques. Does the severity of their prior problems influence the reasons people give for changing their drinking? There is some indication that respondents who had severe problems prior to remission are more likely to provide negative consequence reasons for change while those with less severe drinking problems describe reasons that correspond to maturing or drifting out of a drinking lifestyle [7]. What do reasons for change mean? Can current or former heavy drinkers have true insight into their reasons for change (or, for that matter, people's reasons for any change)? Are retrospectively recalled reasons for change more or less likely to be ‘biased’ in some way as opposed to prospectively endorsed reasons? Some researchers have taken to discussing the reasons that people give for change as narratives or stories [8, 9]. Further, they have pointed out that the absolute accuracy of respondents’ reasons for change may not be as important as the use to which respondents put these reasons. The ‘story’ that former heavy drinkers have about why they quit could be the ongoing linchpin that they use to motivate sustained remission. What of the Matzger et al. study [1]? This well-conducted research makes some acknowledgement of the potential biases associated with the particular reasons for change they found. The combination of a retrospective recall of reasons for change combined with a prospective analysis of how these reasons are related to sustained remission could be regarded as a good test of what reasons (or stories) might work well as motivating factors in maintaining remission. The point of this commentary is not that any one of the methods used in reasons for change research is necessarily better or worse than any other, simply that the reasons for change identified in each study need to be considered within the framework of the methods used to collect the findings.
Source: Addiction
Author: JOHN CUNNINGHAM
Institution: Centre for Addiction and Mental Health
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W1700645869