In Stories We Trust: Studies of the Validity of Autobiographies

There is a dearth of research addressing the validity of life stories, or autobiographies. In part, this is because it is unclear how such data sources might be validated. This article recommends two different perspectives for obtaining evidence relevant to the validity of autobiographical data. A study is conducted from each of these perspectives, and multiple validity estimates are obtained in each investigation. The construct validity of standard psychological constructs (e.g., assertiveness, trustworthiness) obtained in Study 1 from autobiographies was equal to that of standard instruments designed to assess these constructs. Evidence for the validity of life themes, extracted from autobiographies in Study 2, was also apparent. Because life histories, autobiographies, and case studies are relied on heavily in the practice of counseling psychology and are rapidly regaining popularity in scientific studies, these positive findings suggest the wisdom in this increasing use of life history data. The use of autobiographical materials in psychological research has a long history (Allport, 1942; Murray, 1938; White, 1952). Unfortunately after a golden age of life history research (spanning the 1930s and early 1940s), studies of life stories generally declined for several decades (from the end of World War II into the early 1970s). However, a vigorous renaissance of research on life stories, autobiographies, and psychobiographies has occurred in the past decade and a half (Cocks & Crosby, 1987; Howard, 1989, 1991; McAdams, 1985, 1990; McAdams & Ochberg, 1988; Rabin, Aronoff, Barclay, & Zucker, 1981; Runyan, 1982; H O W A R D E T A L . , J O U R N A L O F C O U N S E L I N G P S Y C H O L O G Y 3 9 (1 9 92 ) 2 Spence, 1982; Tyler, 1978). (See McAdams, 1985, pp. 19–24, for a more detailed history of research in personology.) On the other hand, life stories have always been a staple of psychological practice. For example, therapy usually begins with an invitation to a client to tell his or her story. Counselors have favored ways of phrasing their readiness to hear the client’s tale, such as “Can you tell me what brings you here?” “How can I be of help to you?” or “What seems to be the problem?” Clients generally know that these invitations do not request the telling of one’s complete life story. Rather, they understand that their task is to tell the part of their life story that appears most relevant to their presenting problem. Whereas the life stories of clients often represent the staple data of counseling practice (whether of the personalemotional or career-vocational sort), life stories have been underutilized in counseling psychology research. If one could demonstrate the validity, viability, or trustworthiness of autobiographical data in some way, it might help to integrate the science and practice functions in the discipline of counseling psychology (see Heppner et al., in press; Howard, 1985, 1986). With the continuing importance of life stories in counseling practice, and with the resurgent importance of life stories in research, this might be an appropriate point at which to consider the evidence that tests the validity of life stories. Are there ways to transform current validation strategies that will enable researchers to appreciate the validity of life themes extracted from autobiographies? That is, How can validation strategies be modified to obtain empirical evidence for testing the validity of stories (such as life histories, case studies, etc.) in which counselors often place great trust? There has always been concern about the validity of life histories (e.g., Gottschalk, Kluckholn, & Angell, 1945; Murray, 1938; Runyan, 1982, 1990), but the record of empirical validation efforts is sparse. Allport, Bruner, and Jandorf (1953) studied life histories of 90 persons who lived through the Nazi experience in Germany. A subset of 10 subjects were interviewed by one of the investigators, and these interview summaries were compared by independent judges to summaries extracted from the life histories. The investigators reported a 95% agreement rate for summaries of the same person from the two independent rating sources. Unfortunately, the report of this research is presented in one brief paragraph, in no greater detail than is described here. Dailey (1971, chapter 6) showed that judges could use fragments of life stories to make predictions about subjects’ later life events with greater accuracy than would be expected by chance. An extensive literature search revealed a dearth of literature purporting to assess the validity of life histories (see Allport, 1942, and Rubin, 1986, for exceptions). However, the absence of such a literature is understandable, given the histories of the fields of personality, personology, and measurement theory. Personology (i.e., research focused on the study of whole lives; cf. Murray, 1938) developed as an outlier, in contrast to the mainstream research in personality, which focuses on personality traits. The scientific study of the whole person, however, was never well integrated into the mainstream of American psychology in general, nor personality psychology in particular. . . . Personality psychologists became experts in “extraversion” or H O W A R D E T A L . , J O U R N A L O F C O U N S E L I N G P S Y C H O L O G Y 3 9 (1 9 92 ) 3 the “need for achievement,” but few deemed it worthwhile or profitable to become experts on “persons” (see the reviews provided by Adelson, 1969; Block, 1981; Carlson, 1971, 1975; Helson & Mitchell, 1978; Maddi, 1982; Singer & Singer, 1972). (McAdams, 1985, p. 21) The split of personology from mainstream personality research was also important because greatly differing notions developed for testing the validity of traits (in personality theory) and the validity of life themes (in personology theory). Of equal importance for the purposes of this article, the two intellectual traditions developed differing notions of the kinds of evidence that might be marshaled in support of the validity of various measurement approaches (self-report questionnaires, behavioral measures, autobiographies, etc.). Finally, because of this split, applied psychologists often have been actively dissuaded from conducting research on intensive case studies. For example, the participants at the Boulder Conference, Clinical Training in Psychology, Boulder, Colorado, 1948, doubted that a single case study, no matter how brilliantly executed, could satisfy the requirements of a doctoral dissertation but noted, “However, this question deserves further study before a categorical denial of the suitability of this form of research is made” (Raimy, 1950, p. 89). Mainstream personality theory developed in conjunction with standard measurement theory (Cook & Campbell, 1979; Cronbach, 1971). The concepts of face validity, content validity, criterion validity, and construct validity became central to both fields and are in fact familiar to most students of psychology. The ways in which such concepts relate to the validation of psychological traits is well documented (e.g., Cook & Campbell, 1979; Cronbach & Meehl, 1955). Yet, how can these concepts be applied to biographical data? One can well imagine the ways in which more rudimentary forms of measurement validity (i.e., face and content validity) might be determined for life stories. What is less clear is how the more compelling forms of measurement adequacy (i.e., criterion and construct validity) relate to life history data.