sign in sign up
<<
Home , Personality, Personality psychology, Psychology

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Leadership Institute Faculty Publications Institute

2006

Using Multiple Methods in

Brent W. Roberts University of Illinois at Urbana-Champaign, [email protected]

Peter D. Harms University of Nebraska - Lincoln, [email protected]

Jennifer L. Smith University of Illinois at Urbana-Champaign

Dustin Wood University of Illinois at Urbana-Champaign

Michelle Webb University of Illinois at Urbana-Champaign

Follow this and additional works at: https://digitalcommons.unl.edu/leadershipfacpub

Part of the and Quantitative Methods Commons

Roberts, Brent W.; Harms, Peter D.; Smith, Jennifer L.; Wood, Dustin; and Webb, Michelle, "Using Multiple Methods in " (2006). Leadership Institute Faculty Publications. 24. https://digitalcommons.unl.edu/leadershipfacpub/24

This Article is brought to you for free and open access by the Leadership Institute at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Leadership Institute Faculty Publications by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Published (as Chapter 22) in Michael Eid and , eds., Handbook of Multimethod Measurement in Psychology (Washington, D.C.: American Psychological , 2006), pp. 321–335. Copyright © 2006 American Psychological Association. Used by permission. “This article may not exactly replicate the final version published in the APA edition. It is not the copy of record.”

Using Multiple Methods in Personality Psychology

Brent W. Roberts, Peter Harms, Jennifer L. Smith, Dustin Wood, and Michelle Webb University of Illinois at Urbana-Champaign

In many ways, Campbell and Fiske’s (1959) arti- adequate, and should be avoided at all costs. Or, cle on multitrait-multimethod (MTMM) approaches conversely, we should all be doing multimethod to validity has stood like a Platonic studies. This directive fails to consider the empiri- for personality since its publication. In cal I fact just mentioned, which is that measures of the ideal study, and scientific world, our constructs the same construct seldom correlate highly enough should converge in a robust and coherent fashion across methods to warrant averaging across meth- across diverse methods. Moreover, we should all as- ods (Fiske, 1971). What are needed, rather than pire to use multiple methods in both validating our mandates to perform multimethod studies, are the- measures and in investigating our ideas. Interest- oretical models that successfully incorporate and ingly, that Platonic ideal is not realized as often as explain both the overlap and lack thereof of identi- expected. If one looks closely at the empirical litera- cal constructs across methods. In our following re- ture in personality psychology, the expectation that view, we attempt to highlight the few abstract constructs should converge across methods and empirical examples that have done so. is seldom met at the level implied in the original ar- Our third point is that the focus on multiple ticle. This is not to argue that the Platonic ideal is methods has inadvertently led to a misguided not appropriate. Rather, one of the major points we boondoggle to search for the methodological holy would like to make in this chapter is that the ideal of grail—the one method that deserves our ultimate the MTMM approach is often taken too literally and . Campbell and Fiske (1959) should not be is sometimes misused or misinterpreted. Why speak saddled with full responsibility for this phenome- such apostasies? In large part, because we are mo- non beyond the fact that they made it clear that we tivated to reiterate points made, ironically, by Fiske should be pursuing multiple methods. Leave it to himself (Fiske, 1971). nature that psychologists would take that What are these points? The first is that different idea and try to one up the multimethod approach methods, or modes as Fiske (1971) described them, by finding the ultimate method. Thus, we have are seldom innocuous. Thus, the literal assump- had hyperbolic statements made for and against tion drawn from Campbell and Fiske (1959) that particular methods made since the 1960s. People measures of similar constructs drawn from differ- have argued that -reports are fundamentally ent methods should converge quite robustly is not flawed and indistinguishable from response styles met as often as we would like. This can lead to er- (Hogan & Nicholson, 1988; Rorer, 1965), that ob- roneous and nihilistic conclusions, such as the con- server ratings are the seat of personality psychol- struct of interest, like , does not exist ogy (Hofstee, 1994), that projective tests do not (e.g., Lewis, 1999). The second point is the assump- work (Dawes, 1994), and that we should prioritize tion that monomethod studies are problematic, in- online measures over all other techniques (Kahn-

321 322 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) eman, 1999). As will be seen in the following re- Much attention has been dedicated to finding a views, none of these positions is defensible. working taxonomy of traits, and many accept the As the methods used are often tied inextricably Big Five as a minimal number of domains (Gold- to the ideas in a field, we will first provide a work- berg, 1993). We prefer the Big Seven (Benet-Mar- ing definition of the field of personality psychology tinez & Waller, 1997). The Big Seven adds global that will serve as an organizing for the positive and negative evaluation to the Big Five subsequent review. As will be seen, this is a true and is a better representation of the entire trait do- case of form following function, as the content cate- main. We prefer this model because, as will be seen gories within the field of personality are each dom- later, one distinct characteristic of our definition of inated by specific methods. Then, we review recent personality is the inclusion of reputation as a key multimethod studies within and across the con- element that has been underemphasized in the tent domains of personality psychology. We will field. And although people may not describe them- end with some about particulars of multi- often with terms such as “evil” or “stun- method approaches in personality psychology. ning,” they do describe others in these terms. , broadly construed, is the second do- main of personality and subsumes all the things WHAT IS PERSONALITY PSYCHOLOGY? that people feel are desirable. We define the domain of motives as what people desire, , and strive Personality psychology is the study of the indi- for-or perhaps more simply, what people want to vidual differences in traits, motives, abilities, and do. This category includes values, interests, prefer- life stories that make each unique (Rob- ences, and goals (e.g., Holland, 1997), in addition to erts & Wood, in press). Figure 1 depicts the pri- the classic notion of motives and (e.g., Mur- mary units of focus in our definition of personality, ray, 1938). Currently, this domain is less coherent which reflects what we describe as the neosocioan- than the trait domain because there is no working alytic perspective on personality. For the purposes taxonomy to organize the units of analysis. None- of this chapter, we will focus on the left-hand por- theless, there are striking commonalities across di- tion of the model and forgo a discussion of social verse areas, such as motives, goals, values, and in- roles and culture, so as to focus on the traditional terests. For example, in each of these domains of content and methods of personality psychology. motivation, one can find superordinate themes of As can be seen in Figure 1 there are four “units of (desire for status and power) and commu- analysis” or domains that make up the core of per- nion (desire for and belonging). So, for sonality: traits, motives, abilities, and narratives. example, the primary motivational units have been These four domains are intended to subsume most, achievement, power (agency) and affiliation (com- if not all, of the broad categories of individual dif- munion; Smith, 1992). The -order factors that ferences in personality psychology. subsume most dimensions also reflect power The first domain, traits, subsumes the enduring and affiliation (Schwarz & Bless, 1992). patterns of thoughts, , and that The third domain reflects abilities and the hier- distinguish people from one another. Or, more eu- archical models identified in achievement litera- phemistically speaking, traits refer to what people tures-that is what people can do (Lubinski, 2000). typically think, feel, or do. In this regard, we view Specifically, is an individual’s “entire traits from a neo-Allportian perspective (Funder, repertoire of acquired skills, knowledge, learning 1991). From this perspective, traits are real, not fic- sets, and generalization tendencies considered in- tions of people’s semantic . They are causal, tellectual in nature that [is] available at any one not just summaries of . Moreover, they are period of time” (Humphreys, 1984, p. 243). Two learned. Even with identical genetically determined models of abilities prevail. The first decomposes , two may manifest dif- generalized intelligence (g), into constituent el- ferent traits because of their unique life experiences. ements of verbal, quantitative, and spatial abili- Using Multiple Methods in Personality Psychology 323

Figure 1. A neo-socioanalytic topographical model of personality psychology.

ties. The second decomposes g into two domains a simple classification on these charac- of fluid and crystallized intelligence (Horn & Cat- teristics, and one common strategy is the use of il- tell, 1966). The most radical feature of our system lustrative stories (McAdams, 1993) or scripts (de St. is that individual differences in ability should be Aubin, 1999). People find it very easy to tell stories a primary focus of personality researchers. How about themselves, others, and their environments. people differ on abilities is clearly important from These narratives in turn help people create mean- both pragmatic and theoretical perspectives, and ing and purpose in their lives and, predictability, any description of an individual life would be in- in the events they observe and experience and pro- adequate if it were not included. vide explanations of how people have come to be The final domain focuses on the content of per- in their present circumstances. sonal stories and narratives that people use to un- The identification of these four domains is cur- derstand themselves and their environments (Mc- sory and deserves greater attention. Nonetheless, Adams, 1993). A critical point to consider in any we feel that this is a sufficient start to organiz- model of personality is that although individu- ing the units of analysis found within personality als can be classified in terms of traits, abilities, and psychology and, more clearly than other systems, goals, they often (if not generally) communicate in- identifies what we study and, in part, the methods formation about themselves quite differently than we use to study individuals. 324 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006)

Personality Is a Multilevel Phenomenon two psychological media: the and the rep- utation. Identity reflects the sum total of opinions A key component of our neosocioanalytic per- that are cognitively available to a person across the spective on personality is that the domains of traits, four units of analysis described earlier. We use the motives, abilities, and narratives can be differenti- term identity for several . The most impor- ated in hierarchical terms (see Hooker, 2002; Hooker tant is the fact that identity pertains to both & McAdams, 2003; Mayer, 1995; Roberts & Pomer- the content of self- and the meta-cog- antz, in press). For example, at the broadest level nitive of those same self-perceptions. of the trait domain one finds the personality traits Specifically, people can simultaneously see them- found in standard omnibus personality invento- selves as “outgoing” and a “carpenter” and feel ries. These are often the traits that make up the now more or less confident about those self-perceptions. ubiquitous measures of the Big Five. The midlevel Or, people can see themselves as agreeable (self- of the continuum can be conceptualized by narrow percept) and at the same time see their agreeable- traits, such as the subfacets of the Big Five (Rob- ness as changeable or not (meta-cognitive percept). erts, Bogg, Walton, Chernyshenko, & Stark, 2004). These latter metacognitive aspects of identity, re- These constructs are broader than discrete behav- flected in constructs such as entity versus- incre iors but less broad than traits, as they are often con- mental orientation (Dweck & Leggett, 1988), iden- strained to specific roles and interpersonal contexts tity achievement, identity clarity, and importance, (e.g., relationships, work, and friendships). Pre- a significant role in personality assessment sumably, these midlevel constructs are more stable and development (Roberts & Caspi, 2003). than discrete behaviors and less stable than broad Reputation is others’ perspectives about a per- traits (e.g., Conley, 1984). At the most narrow level, son’s traits, motives, abilities, and narratives we find the constituent elements of traits: thoughts, (Craik, 1993). There is a tendency to consider ob- feelings, and behaviors. So, for example, one can be server ratings, or in this case, reputational ratings, a depressed person, indicating a broad generaliz- as higher quality than self-reports (Hofstee, 1994). able pattern of depressed across time and situ- This position holds some merit, as a good assess- ation, yet experience different daily moods or states ment of a person’s reputation entails asking the that do not correspond directly to one’s trait level. opinion of more than one person. Thus, reputa- The hierarchical structuring of each domain of tional ratings, by their very nature, are often in- personality adds another layer of methods on top trinsically more reliable than self-reports because of the methods typically identified within person- self-reports only entail the opinion of one person. ality psychology (see following). So, not only can Reputations also guide significant decisions, such one assess personality through global ratings of as whether to hire a person, admit them to gradu- personality traits, but also through daily - ate , marry them, or simply be their friend. ings or frequencies of behaviors. Or, similarly, one From our perspective, the self-reports used to could assess a person’s through broad assess identity and the observer ratings used to as- ratings of values and interests or the relevant ac- sess reputation both afford unique, yet flawed, in- tions they take in their lives, such as exercising and formation about a person. Certain psychological well as, manifestations of valuing . phenomena, such as feelings of , may best The information gleaned from these different lev- be accessed through self-reports of identity On els constitutes different methods that are partially the other hand, determining a person’s true levels overlapping, yet distinct in important ways. of might be better assessed through the opinion of their friends and relatives who The Methodological and Conceptual Fulcrum: may be less defensive about another person’s be- Identity and Reputation havior than their own. Each perspective is poten- According to our conceptualization of personal- tially defective, in that neither the persons report- ity, the components of personality are manifest in ing on themselves nor the persons reporting on a Using Multiple Methods in Personality Psychology 325 friend or relative are perfectly accurate. Accepting between self and observers across psychological the fact that there are two flawed and distinct ways domains. to understand a person confronts and solves sev- The RAM model has implications beyond the re- eral dilemmas that have plagued personality psy- lationship between observer/reputation and self- chology for decades. For example, it automatically reports/identity. It also applies to the accuracy of incorporates the fact that people can and do at- self-reports themselves, in the absence of any ob- tempt to manage their identity to shape their rep- server data. For example, we often ask young peo- utation. People do not always tell the whole truth ple to rate themselves on a variety of personality about themselves to employers, friends, , dimensions without ever asking ourselves whether and strangers. Self-presentation is a fact in human these individuals make good judges of their own nature and must be successfully incorporated into personality. For example, a young person may any of personality and cannot be incorpo- be more than willing to say that they are a good rated without a distinction between identity and leader, based not on experience but on the reputation (Hogan & Roberts, 2000). that someday they will become one (relevance). Or, Despite the spirit of the MTMM approach quite possibly, a person may do something rele- spelled out by Campbell and Fiske (1959), the con- vant to a trait but not notice it (detection). That is vergence of self-reports and observer ratings of to say, people may not be aware of the importance personality and other phenomena has never been or relevance of the diagnostic nature of their own as high as one would hope. In most cases, the con- behavior. Finally, people may use self-relevant in- vergence averages between .3 and .6 (Funder, formation in idiosyncratic ways that might not 1987). One of the clear conceptual and method- conform to how scientists define or understand a ological advances in the field of personality psy- nomothetically derived construct. With the excep- chology is the Realistic Accuracy Model (RAM; tion of the availability stage of the RAM model, it Funder, 1995), which provides a clear theoreti- seems that the remaining mechanisms for accuracy cal model identifying why identity and reputa- can be applied to a number of issues across psy- tion are not more highly correlated. In this model, chology and personality psychology in particular. for a strong tie to exist between self-reports and Measures of identity and reputation also do not observer ratings four conditions must hold. First, correlate as high as expected because they are as- the person being perceived must do something sessed through distinct methods that afford dif- relevant to the psychological dimension of inter- ferent types of information (see Meyer et al., est. If one wants to judge whether a person is con- 2001). Clearly, identity-related assessments per- scientious or not, then it is imperative that they mit greater access to internal states and experi- act in a conscientious fashion. Second, the behav- ences that do not happen or are not visible in the ior, , or must be displayed in a company of others. Reputations, on the other hand, way that it is made available to the observer. Like may be less tarnished with self-enhancement ten- the proverbial tree falling in an empty forest, pri- dencies and provide a more objective profile of vate actions do little to influence one’s reputation, the information that is publicly available to peo- unless of course they are made public. Third, the ple or experts (Hofstee, 1994). Reputational infor- observer must detect the behavior. If the person mation may not be ideal because its validity is un- watching does not perceive the behavior, then it dermined by the fact that observers do not have might as well not have occurred. Finally, the ob- complete access to a person’s thoughts, feelings, served act must be used in an appropriate way. and behaviors (Spain, Eaton, & Funder, 2000), al- For example, to some people, being clean may be though conversely, individuals may be unaware of a sign of , whereas to others it some of their own behavioral tendencies that im- may be an indication of (e.g., obses- pact their reputations. Using both identity and rep- siveness). The extent to which these four condi- utational information and their tions hold determines the level of correspondence relationship is paramount for the of per- 326 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) sonality. We will find that the distinction between biographical documents to understand individ- identity and reputation runs through each of the ual differences (e.g., L data). The field of person- domains of personality psychology and often acts ality typically utilizes diverse methods in an at- as a fulcrum for understanding multimethod stud- tempt to understand how individuals differ from ies in personality psychology. one another. This also makes studies that combine The methods found within the categories of iden- assessments from each of these disparate domains tity and reputation can be further divided into the intrinsically multimethod studies. We highlight ex- of methods that have historically dominated amples of these types of studies from each domain. the field of personality psychology. Broadly speak- There are more traditional multimethod ap- ing, methods of assessment in personality psychol- proaches within each domain of personality. For ogy can be organized around the acronym “LOTS” example, within the domain of personality traits, (Block, 1993). L stands for life data, or the narrative evaluating the of self-reports and observer content of a person’s life. O stands for observer data, ratings has been a constant struggle for several de- which can come from peers as well as trained pro- cades. Within the motives domain a long-standing fessionals. T stands for data and typically re- controversy has been whether to assess motives us- flects objective performance measures. And finally, ing implicit or explicit techniques. We will high- S stands for self-reports, or the subjective inferences light studies within each domain that have endeav- we have about ourselves. Typically, S and L data are ored to use more than one method within domain. acquired through self-report techniques of ratings or interviews. T and O data are acquired through Multiple Methods and Personality Traits observer reports because the tests typically have to As we noted, one of the persistent disputes in be scored by computer or person, and observer rat- personality psychology is between those who be- ings clearly must be acquired through peers, family lieve that self-reports or observer methods should members, or interviewers. These four approaches hold priority in the field. The programmatic efforts to assessment subsume the majority of the method- of David Funder and his colleagues demonstrate ological efforts in personality psychology. that multiple methods bring multiple perspectives to our efforts to understand the behavioral mani- MULTIPLE METHODS WITHIN THE festation of personality traits. For example, people NEOSOCIOANALYTIC FRAMEWORK judging the behaviors of others perceive different OF PERSONALITY cues as more relevant to personality than the indi- viduals themselves (Funder & Sneed, 1993). A more complete conceptualization of personal- In other studies, the usefulness of self- and ob- ity psychology points to many ways in which mul- server ratings of personality have been tested across tiple methods can be brought to bear on the study a variety of domains, including predicting behavior, of personality. First, within each domain of traits, , and personal negativity. The key to test- motives, abilities, and narratives there are rich ing the utility of different methods is separating the methodological and differences. So, for perspectives of self and observer from the criterion example, traits have often been assessed using self- of interest. To do this, Kolar, Funder, and Colvin reports of typical behaviors. Similarly, motives and (1996) set up a study in which the participants pro- goals have been assessed from the perspective of vided self-report personality ratings, close acquain- the person (e.g., S data) and the who tances provided an additional set of personality rat- interprets a such as the TAT (e.g., ings, and the behavior of participants was coded O data). Cognitive ability has been traditionally from videotaped interactions. Thus, the two sets of assessed through tests of maximal performance predictors and criteria did not suffer from meth- (e.g., T data), but can also be assessed via self-re- odological overlap. For behavior in a typical social ports. In contrast, the narrative approach focuses setting, such as meeting a stranger or having a dis- on open-ended interviews, written responses, or cussion, observer ratings tended to predict behav- Using Multiple Methods in Personality Psychology 327 ior better than self-reports (Kolar, Funder, & Colvin, about a ’s depression may not result in high 1996). For example, the correlation between self-re- agreement with the child’s assessment (e.g., Lewis, ported tendency to initiate humor and actual behav- 1999). Rather than seeing this as an indictment of ior of initiating humor was .09. In contrast, a com- either perspective or the construct of interest, we posite of the rating of the participants’ tendency to should use both of these perspectives and more initiate humor of two close acquaintances correlated (e.g., teachers, peers, ) to gauge the nature .23 (p < .05) with actual behavior. Clearly, what we and progress of the phenomenon. For example, a believe to be a is not perceived by others to be child may have effectively hidden depression from funny, which might explain why more people don’t his or her parents, but not hidden the same phe- laugh at our . nomena from his or her peers. The discrepancy it- Spain et al. (2000) used a similar design to both self may be both interesting and relevant to the replicate Kolar et al. (1996) and extend the design to experience of depression, as it might reflect alien- see if self-reports might be superior in specific set- ation and disengagement from parents that might tings, such as when one is predicting rather be a contributing factor to the depression. than behavior. Consistent with expectations, self-re- The perspective that no single method holds pri- ported personality ratings were more strongly re- ority extends to arguments against the use of pro- lated to experience sampling assessments of emo- jective measures (Dawes, 1994). For example, in tion than observer ratings of personality traits. This our meta-analysis of the longitudinal consistency presumably derives from the fact that emotions are of personality trait measures (Roberts & DelVec- internal events that are not always shared with oth- chio, 2000), we found that projective measures of ers as overt, visible behaviors. Their private nature personality traits were as consistent as observer makes them a natural target for self-reports rather and self-report methods of personality assessment. than observer ratings. Interestingly, self-reported Moreover, in particular cases, projective measures personality ratings did better than observer ratings outperform other methods, such as in the assess- of personality in predicting social interactions. For ment of dependency (Bornstein, 1999). This does example, self-reported extraversion was correlated not to provide a ringing endorsement for projec- with demonstrating social skills, as judged by a set tive tests, as it is clear that specific projective tests of trained raters, whereas a composite of acquain- and particular measures derived from projective tance ratings was essentially uncorrelated with the tests do not demonstrate adequate and same behavior. In fact, for extraversion, self-reports validity (Lilienfeld, Wood, & Garb, 2000). None- were twice as good as observer ratings of extraver- theless, blanket statements that they should not be sion in predicting behaviors. used are not warranted given the evidence. Clearly, based on this research alone, we cannot The idea that perspectives that differ in terms of make any strong generalizations about the supe- their hierarchical relationship to personality pro- riority of self-reports and observer ratings of per- vide different, yet equally valid information was sonality This is itself important, as it undermines demonstrated nicely by a recent study of satisfac- claims that any given perspective is superior. Stud- tion with one’s vacation (Wirtz, Kruger, Napa- ies that actually use multiple methods arrive at Scollon, & Diener, 2003). In this study, participants more equivocal conclusions. This point is driven rated how satisfied they thought they would be home conceptually in a review of the utility of psy- with an upcoming vacation. In addition, they com- chological assessment (Meyer et al., 2001). In de- pleted an online assessment of their emotional ex- scribing the importance of using multiple meth- periences during the vacation using experience- ods of personality assessment in clinical settings, sampling methods. A week later, they rated how Meyer et al. (2001) argued that each method af- satisfied they were with their vacation. Interest- fords a clinician, and by default a researcher, infor- ingly, anticipated and retrospective ratings of - mation that may not be strongly overlapping, yet isfaction were much higher than online ratings of still quite valid. That is to say that asking parents satisfaction, indicating a slight disjoint between 328 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) actual experience and higher-order evaluations tivation is not accessible through conscious pro- of that same experience. Moreover, the different cesses and that it should be interpreted through methods yielded different information. The online expert analysis of material generated by a person experiences were strong predictors of the retro- without their knowledge of what is being assessed. spective ratings of satisfaction, which were in turn In contrast, goal theorists have no qualms about as- the most important predictor of wanting to go on a sessing goals using conscious processing. similar vacation in the future. The effect of actual The need approach to motivation is clearly con- experience on the desire to go on a similar vacation nected historically to the use of the Thematic Ap- was entirely mediated by the higher-order gener- perception Test (TAT), which was initially devel- alizations about satisfaction, which indicates that oped by Murray (1938). Following the belief that the different methods yielded complementary in- individuals are unaware of their motives and un- formation rather than redundant information. This able to report accurately on them, the TAT was de- study counters the argument that online assess- signed as a projective technique under the belief ments should be prioritized over broader, some- that “when a person interprets an ambiguous so- times retrospective reports of personality (e.g., cial situation he is apt to expose his own personal- Kahneman, 1999), as it was the global self-reports ity as much as the phenomenon to which he is at- that predicted long-term intentions rather than di- tending” (p. 531). These together form rect, behavioral measures of experience. the theoretical basis of the TAT, where participants Within the trait domain, we find many of the are asked to take the part of story-writers and cre- classic arguments about multiple methods, such ate stories on the basis of ambiguous pictures. Al- as the utility of self-reports versus other ratings though the traditional TAT is the one and newer perspectives manifest in assessing per- most commonly associated with the assessment of sonality across multiple levels of breadth or across Murray’s needs, several alternative routes to the as- different contexts. Consistent with the neosocio- sessment of Murray’s needs have been developed. analytic framework that a differentiated conceptu- For instance, Schmalt (1999) developed a “semi- alization of personality leads to a multimethod ap- projective” grid technique in which individuals are proach, each of these different methods revealed asked to rate what characters in TAT-like pictures complementary and useful information. What we are thinking or feeling from a fixed set of options. still lack, of course, are theoretical systems to ac- The second school of thought within the domain count for the complementary rather than overlap- of motivation is that of the goal approach, which ping nature of the information gleaned from dif- begins by asking individuals what they are typi- ferent methods. Systems like the RAM model are cally trying to do in their everyday lives. Whereas a step in the right direction, but more conceptual theorists working within the need approach to mo- and theoretical work is needed. tivation state that behavior is determined largely by discrepancies between actual states and uncon- Motives and Goals scious motives, goal theorists believe behavior is largely influenced by discrepancies that are con- Research in the domain of motives has had two sciously accessible (Emmons, 1986, 1989). Further, major methodological and theoretical , whereas needs are conceptualized as broad, de- which address the study of this broader question contextualized, and fundamental constructs (Win- in quite different ways. The first school, the need ter, John, Stewart, Klohnen, & Duncan, 1998), goals approach to motivation, begins with the assump- are assumed to vary hierarchically in their level of tion that people are often unaware of the funda- abstraction, ranging from specific and short-term mental forces that motivate their behavior. The sec- goals such as “what I’m currently concerned with ond major school, the goal approach, attempts to doing” (Klinger, 1975; Little, 1983), to more-endur- understand explicit motives and interests as the ing midlevel constructs such as personal strivings means to reach a deeper, underlying understand- that reflect “what I’m typically trying to do” (Em- ing of motivation. Need theorists believe that mo- Using Multiple Methods in Personality Psychology 329 mons, 1986) and finally to more broad and long- Recently, research has attempted to form a more term life goals such as establishing a career or find- complete picture of the associations between motive ing a relationship partner (Roberts & Robins, 2000). measures by looking simultaneously at the mea- Each of these levels is associated with a slightly dif- sures used by need and goal theorists. Emmons and ferent method, although generally these methods McAdams (1991) examined the relations between are idiographic, allowing respondents to give open- the Jackson PRF, personal strivings, and TAT mea- ended responses to the instructions and rely on sures for the assessment of the achievement, affili- conscious acknowledgement of one’s aspirations. ation, intimacy, and power motivations. The au- Consistent with Fiske’s (1971) argument that thors found modest relations between matching method can have a profound effect on construct TAT and striving categories for achievement, inti- validity, one of the long-standing controversies macy, and power motives, indicating that, to some within the field of motivation is whether implicit extent, these methods may be measuring the same or explicit methods of assessing motivations as- underlying construct. On the other hand, the self-re- sess the same constructs. McClelland, Koestner, ported PRF was related to matching dimensions of and Weinberger (1989) argued that measures such personal strivings for power and achievement mea- as the TAT and questionnaire measures such as the sures, but was irregularly related to the TAT mo- Jackson PRF are measures of distinct constructs, la- tives. For instance, self-reported dominance was beled implicit needs and self-attributed needs, re- related positively to TAT achievement, but was un- spectively. Presumably, implicit measures should related to TAT power. The authors concluded that predict operant behaviors that are relatively un- personal strivings may lie somewhere between self- controlled by the environmental, such as job level attributed motives and implicit motives in that striv- attained in and behavior occurring ings appeared to relate to the TAT and PRF better under natural conditions (McClelland, 1980). Self- than these scales relate to each other. attributed motives should be more predictive of re- However, a second study looking at motives for spondent behaviors, such as school grades, person- power, affiliation, and achievement (King, 1995) ality, and intelligence tests, where the behavior is failed to replicate Emmons and McAdams’ (1991) elicited and constrained by environmental stimuli. findings. King (1995) failed to find direct relation- In a meta-analysis of the literature on achieve- ships between the TAT and a battery of other mo- ment motivation comparing the utility of self-at- tive measures, including the PRF as well as strivings, tributed ratings of motives to implicit measures reported wishes, and early coded us- of motives, Spangler (1992) found support for this ing Winter’s (1991) running text system. This study . Implicit measures were more predic- also failed to find relationships between the PRF and tive of outcomes when attaining the outcomes in- strivings measures of power or affiliation motives. volved challenges or incentives that were intrinsic The lack of relationship between the PRF and TAT to the task, such as moderate risk and time pres- motives conformed well to Spangler’s (1992) finding sures, whereas self-attributed measures were more of an average correlation between TAT and self-at- predictive of performance in tasks that involved tributed motives of r = .09 across 36 studies, which social incentives, such as challenging goals set up suggests that implicit and self-attributed motives are by the experimenter or norms that encouraged not independent, but are very nearly so. Although achievement. Interestingly, implicit motives were clearly more research needs to be done, it seems rea- also found to decrease in their relation to task per- sonable to conclude from these studies that the de- formance when the number of social incentives in- gree of commonality between implicit and explicit volved with the task was high. Consistent with the methods of assessing motives is not high. interpretation of implicit needs as somewhat akin The controversy over implicit and self-attributed to intrinsic motivation, Spangler suggested that so- needs has fostered an environment in which very cial incentives may conflict or otherwise suppress few researchers have combined motive measures the effect of implicit needs on performance. with personality measures from the other three do- 330 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) mains of personality (i.e., trait, ability, or narrative). popularity (Boyatzis, 1973). Given the current theo- In their attempt to integrate the domains of traits rizing concerning trait-motive relationships, this can and motives, Winter et al. (1998) suggested that im- now be understood as a by-product of an achieve- plicit motives and personality traits generally inter- ment-extraversion interaction, which explains how act in their prediction of life outcomes. More spe- extraverts and introverts differentially handle their cifically, they hypothesized that motives represent level of achievement motivation (Winter et al., 1998). a person’s fundamental goals and desires, whereas The domain of motives, much like the domain traits channel the expression of these motives to- of traits, is marked by the use of distinct methods ward specific paths. In looking at extraversion in that do not converge as highly as one would like. combination with affiliation and power motivation In part, this divergence is consistent with the the- in two samples of women, the authors found ex- oretical underpinnings of the two approaches. Re- traverts preferred volunteer work, combined fam- searchers who adopt the implicit motive approach ily and work roles more frequently, and had more are skeptical of cognitive appraisals of needs. In stable romantic relationships-but only if they were contrast, researchers who use the self-attributed also high in affiliation motivation. Similarly, ex- approach find this less problematic. It is clear from traverts rated work relationships as more impor- the studies using these two approaches that they tant than introverts, but only if they also were high both bring independent complementary predictive in power motivation. In some cases, crossover in- variance to the research endeavor. Furthermore, teractions were found, where extraverts had more when combined with methods and constructs from satisfying relationships than introverts when both the trait domain, we find clear predictions of im- were high in affiliation motivation, but the reverse portant life outcomes (Winter et al., 1998). was true when both were low on the motive. Win- ter and his colleagues had hypothesized this last in- Abilities teraction by considering the introverted, low-af- filiation individuals as most effective at working Assessing intelligence has traditionally focused alone and unconcerned about the opinions of oth- on multiple forms of test data (T data). Standard ers, whereas introverted high-affiliation individuals measures of intelligence typically attempt to gather desired friendship and but were ineffec- information on a wide variety of traits considered tive at maintaining it because of their awkwardness to be at the core of general mental ability. How- in interpersonal situations. It is interesting to note ever, numerous efforts have been made to move that for all of the life outcomes the authors investi- beyond traditional assessment approaches. These gated, main effects between traits and motives were have included measures of specific cognitive abili- rare, and the importance of the constructs would be ties, intellectual interests, and self-report measures missed if considered separately. of intelligence. Although some studies exist examining the re- Testing cognitive abilities has traditionally in- lationship between different motive measures, we cluded a variety of measurements and techniques, found surprisingly few studies that have looked at such as problem-solving tasks, assessments of the simultaneous interplay of motive and other per- school performance, information acquisition tasks, sonality constructs in the prediction of other out- as well as matrix problems that require highly ab- comes. We suggest that not examining motives in stract conditional discriminations. The reason for combination with other domains results in a failure the success in tapping general cognitive abilities to fully understand the importance of motives, or using a variety of techniques largely has to do with worse, it may lead to erroneous conclusions about the degree to which general mental ability perme- what motives are and do. For instance, early re- ates all learning, reasoning, and problem-solving search on achievement motivation was stymied for abilities. Further, aggregations of measures of spa- decades by the empirical finding of negative -rela tial skills, verbal reasoning, and quantitative abili- tionships between the motive and variables such as ties measure general mental ability more efficiently Using Multiple Methods in Personality Psychology 331 than aggregations of information items because the tions of ECTs form two general factors, perceptual reasoning problems used in these measures typ- speed and working memory (Ackerman, Beier, & ically capture a greater degree of common-factor Boyle, 2002). These factors are, as a result of aggre- variance associated with g (Gustafsson, 2002). Con- gation, both highly related to general mental abil- sequently, the most popular measures of general ity, with working memory being the more highly re- mental ability include a variety of assessments de- lated to g of the two (Ackerman et al., 2002). signed to tap several broad domains highly related Another approach to measuring general mental to general mental ability, such as verbal, quantita- ability has been to use self-reports of intelligence or tive, reasoning, and visuospatial skills. intellectual engagement (Paulhus & Harms, 2004). The search for alternative methods of measur- This approach has been much maligned by intelli- ing general mental ability more purely has often gence theorists because of the fact that self-report led to the use of elementary cognitive tasks (ECTs) intelligence measures rarely exceed validities of .50 that measure processing speed and working mem- with typical tests of maximal performance of cog- ory (Jensen, 1998). These tasks highlight the hier- nitive ability (Paulhus, Lysy, & Yik, 1998). None- archical nature of intelligence and our earlier point theless, the search for better self-report measures that assessments across different levels of abstrac- has persisted because of the interest in finding a tion typically constitute related but different meth- non-stressful and easily administered technique ods. ECTs have proved to be a popular alterna- for obtaining performance information. tive methodology for measuring general mental One of the more comprehensive and successful ability because such tasks avoid the that may self-report measures of intelligence has been the be introduced in measurement by prior training Typical Intellectual Engagement (TIE) scale devel- and experience. It also is argued that basic cogni- oped by Goff and Ackerman (1992). The premise tive mechanisms underlie all thinking, reasoning, behind this scale is that knowledge is accumulated and decision-making processes, and therefore such over time through effort and motivated engage- mechanisms would be substantially related to gen- ment in learning. It is therefore believed that this eral mental ability (Kyllonen & Christal, 1990). measure will better reflect daily behavior because Interestingly, Carroll’s (1993) analysis of the it constitutes a test of typical intellectual perfor- structure of general mental ability showed that tasks mance. This is distinguished from a test of maxi- measuring reaction time, inspection time, and dis- mal intellectual performance, such as an SAT test, crimination ability were only weakly related to gen- where it can be assumed that the individual is eral mental ability. Indeed, early skepticism regard- bringing their full cognitive resources to bear to ing the efficacy of using such measures to measure succeed and attain a better outcome. general mental ability was the result of such mea- The TIE scale has been instrumental in integrat- sures being used in isolation. However, it has been ing measures of the components of Ackerman’s demonstrated that scores on such experimental PPIK theory, a multimethod approach to under- tasks can be aggregated to form a reasonable rep- standing that integrates in- resentation of general mental ability if enough ex- telligence-as-process, personality, interest, and in- periments are carried out across a variety of cogni- telligence-as-knowledge (Rolfhus & Ackerman, tive task domains (Green, 1978). It has been noted 1999). By assessing each of these domains, Rolf- that correlations between combined reaction time hus and Ackerman attempted to get a better ap- scores from a number of ECTs and general mental proximation of the contribution of each to scores ability approach the size typically seen with psycho- on knowledge and intelligence tests. Participants’ metric power tests (Jensen, 1998). Further, the com- general mental ability was assessed using a com- bined scores from a number of ECTs can be used to posite of verbal, mathematical, and spatial abil- predict upward of 70% of the heritable part of the ities. Their and interests were as- variance in general mental ability For the purposes sessed using standard measures of the Big Five of experimentation, it should be noted that aggrega- personality traits, interests, and typical intellec- 332 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) tual engagement. Subjects also completed a battery awards received were recorded. Path modeling of tests measuring their knowledge in a wide va- demonstrated that although achievement orien- riety of domains including , sciences, tation and dependability made independent, al- , and mechanics. This study demonstrated though small, contributions to supervisory ratings, that a substantial higher-order Knowledge factor the impact of general mental ability on supervisory emerges from of the knowledge do- ratings of job performance was completely medi- mains that accounts for approximately 50% of the ated by job knowledge, which in turn was medi- variance in domain knowledge. Further analyses ated by task proficiency. Further, dependability showed that this general factor was significantly was positively related to job knowledge and nega- correlated with crystallized intelligence, which tively related to disciplinary actions. Achievement was represented by a composite of verbal ability orientation was positively related to the number tests. This suggests that the general knowledge fac- of awards a soldier received. The model - tor is highly related, but not identical, to crystal- strated by this analysis shows that although gen- lized intelligence. These findings also suggest that eral mental ability has a huge impact on job knowl- a substantial part of the variance in knowledge test edge, and job knowledge is substantially related to performance remains to be predicted by more do- task proficiency, it is by no means the largest of the main-specific influences, such as interests and per- contributors to job performance ratings by super- sonality. For instance, Extraversion was shown to visors. Personality factors and outcomes associated be negatively related to all but one of the domain with personality factors also make significant di- knowledge tests, with and rect contributions to supervisory ratings. Typical Intellectual Engagement also demonstrat- There are many different approaches to the study ing significant, positive relationships across the and measurement of general mental ability. The knowledge domains. Measures of interests also most successful approaches, and consequently the proved to be related to domain knowledge scores, most widely used, have used measures from across but were more specific with regard to matching content domains to gain a fuller representation of content domains. Realistic interests were related the cognitive functioning required in reasoning, de- to mechanical knowledge domains, Investigative cision making, and other thought processes. Alter- interests were mostly related to science domains, native approaches such as and Artistic interests were most highly related to techniques using elementary cognitive tasks have knowledge domains that reflected the humanities. proved to be successful as indicators of general Like the domain of motives, one finds that com- mental ability, but only when they are assessed and bining tests of cognitive ability with measures aggregated across modalities, content &mains, and taken from other domains, and thus other meth- tasks. Other alternatives, such as self-report mea- ods, maximizes our ability to predict important sures of intelligence and intellectual interest, have outcomes. One of the best multimethod studies shown promise as indicators of general mental abil- that integrated multiple measures of intelligence, ity, but may be best suited to offering a more inte- knowledge, interests, and personality measures grated picture of how basic processes, work- to real-world performance outcomes was Project ing memory, and personality may be related to A (Campbell, 1985). Borman, White, Pulakos, and real-world outcomes in intellectual functioning. Oppler (1991) analyzed data from 4,362 first-term soldiers in nine U.S. Army jobs. Subjects were as- Life Story Narrative as a Means of Investigating sessed for cognitive ability using the ASVAB, as Personality well as job knowledge, dependability, and achieve- Like the first three domains of personality, the ment orientation measures that were developed for use of narrative methods in multimethod research the study. To assess performance, hands-on pro- is a novel occurrence, yet has thus far been infor- ficiency measures and supervisory ratings were mative to the understanding of individual differ- taken, and the number of disciplinary actions and ences. Qualitative assessments of personality be- Using Multiple Methods in Personality Psychology 333 gin at the most basic level with the and A recent multimethod study provides a clear ex- progress to rigorously assessed structured inter- ample of the utility of the narrative approach. Par- views (McAdams, 1996, 1999). Qualitative data ents of children with Down syndrome were con- is frequently gathered in the form of open-ended tacted through a support group mailing list as well questions concerning a topic of interest to the re- as through area hospitals (King, Scollon, Ramsey, searcher. Consistent with the perspective that each & Williams, 2000). The parents were initially asked domain of personality is arranged hierarchically, self-report questions concerning well-being, ad- qualitative data can be examined at both the micro- ministered a projective test of ego development, and macrolevels. Microlevel assessment is con- and asked to write a story about when they were cerned with specific linguistic patterns within the first told that their child had Down syndrome. Two narrative such as pronoun usage or specific word years after the initial assessment, parents again re- type frequencies (e.g., Pennebaker & Francis, 1996). sponded to self-report measures of well-being and In contrast, macrolevel assessment focuses on the a projective measure of ego development. The nar- broad themes throughout a narrative, such as re- ratives were assessed by three independent raters demption sequences (McAdams, Reynolds, Lewis, for themes of accommodation (exploration, shifts Patten, & Bowman, 2001). Such thematic coding in perspective, activity) and closed (, neg- is often developed by the researchers after 1 lis- ative affect). Parents who were low in ego devel- tening to interviews or reading written narratives. opment at Time 1 who wrote in an accommodative Trained coders can then rate each qualitative da- manner demonstrated increased ego development tum on the varying themes of interest. Topics that 2 years later. Parents who wrote narratives in both are open to narrative methods are limited only by an accommodative and closure style had higher the and ambition of the researcher, and feelings of -related growth at the 2-year fol- the richness of the data can afford multiple oppor- low-up. This research provides an example of how tunities to better understand the personality of an healthy processing (i.e., exploring the impact of the individual. For example, McAdams’ life story in- event on the self and discovering a positive resolu- terview (McAdams, 1996, 1999) asks people to de- tion about this experience) of difficult events on an scribe low points, turning points, and religious be- individual level allows for healthier, more mature liefs among other experiences. Each of these stories functioning later in life (King et al., 2000). can be examined individually for specific types of Helson (1992) examined more general identity experiences (e.g., questioning of parents’ religious threatening events in the writing of women’s dif- beliefs, difficult times) to broad life-span themes, ficult life experiences in an ongoing, multimethod such as agency and communion. Qualitative data . Information was gathered can thus be converted into data that is quantita- about the age at which women experienced diffi- tively assessable without losing the nuances of the cult, identity-changing life events and various per- qualitative form. Additionally, excerpts from qual- sonality factors that influenced the onset of such itative data may be used to reiterate a theoreti- experiences. In addition, information about iden- cal point. Examples from three studies will help tity status (achieved, moratorium, foreclosed, and to illuminate these methodologies. A substantial diffuse) was used to understand the and amount of narrative research concerns the reaction effect of difficult times. Women who had a diffuse of an individual to difficult life events in his or her identity presented more themes related to nega- life. Theoretically, the manner in which an individ- tive evaluations of themselves. Foreclosed women ual responds to traumatic experiences that threaten wrote mainly about having bad partners or over- his or her view of self and the surrounding world load. Achieved/moratorium women wrote primar- is critical for understanding the identity of that in- ily about becoming psychologically independent dividual. If an individual is able to construct a co- and its consequences. Additionally, as women’s herent self from a difficult life event, he or she is vulnerability began to decrease and confidence be- considered to have a healthy identity. gan to increase on personality measures around 30 334 Roberts et al. in Handbook of Multimethod Measurement in Psychology (2006) years of age, an increase in identity themes occurs. portant variance in addition to standard methods, This research suggests that the rewriting of the life such as self-reported personality traits. story occurs in middle age for women, and that this is associated with an increased importance of CONCLUSION independence, which is in turn related to healthy identity functioning. It is clear from our review that the field of per- Pals (2005) combined the narratives of the sonality psychology is intrinsically a multimethod women in the Mills study with themes parallel to field. Within and across each domain of personal- those of King et al. (2000) to illustrate not only the ity, methods as diverse as self-reports, observer rat- correlation between personality on the trait level ings, projective tests, test of maximal performance, and narrative level, but also the dynamic interac- and qualitative interpretations of narratives are tive processes of trait and narrative conceptual- brought to bear on understanding individual dif- izations of personality Narratives from women ferences in thoughts, feelings, and behaviors. Con- who had participated in the longitudinal study sistent with our neosocioanalytic framework, many described earlier (Helson, 1992) were coded for of the methods correlate quite strongly with con- themes of resolution (overall resolution, positive tent. Self-reports and observer ratings tend to be ending, low negative ending, low lasting wounds, used more often in the assessment of personality coherent ending) and impact on self (open re- traits, and the complement of hierarchically related sponse, narrative complexity, low self-, constructs such as affect and behavior. The use of acknowledged wounds, positive self-transforma- projective tests bridges domains, but is primarily tion, and active approach). These two dimensions located in the assessment of motives. Tests of max- were then used in conjunction with age 21 and age imal performance have the potential to bridge do- 52 responses to personality measures of open ver- mains, but are similarly found almost entirely in the sus defensive (a combination of tolerance of content domain of abilities. Finally, one’s story is al- ambiguity and reverse scored ) and ego- most exclusively the domain of methods that fo- resiliency to predict physical and psychological cus on life data. Moreover, within each domain re- health outcomes at age 61. Findings demonstrated searchers are beginning to use multiple methods to that whereas coping openness at age 21 was related assess the hierarchically related constructs within to clinician-rated at age 61, this relation a content category, such as when broad trait mea- was mediated by the extent to which women com- sures are combined with the assessment of daily posed a narrative that was open to expressing the mood or behavior (e.g., Wirtz et al., 2003). impact of the negative events on the self. Further, Despite the impressive methodological plurality whereas a resolved narrative was related to sub- across and within domains in personality psychol- jective well-being at age 61, this effect was medi- ogy, there remains a tremendous unrealized poten- ated by ego resiliency at ages 21 and 52, suggesting tial to bring multiple methods to bear on relevant a dynamic interaction of trait and narrative person- topics. For example, the use of test data in domains ality in relation to healthy functioning. other than abilities remains untapped, despite pro- provides the researcher vocative studies pointing to the potential to assess with an ability to not only examine the individ- personality traits in ways other than asking some- ual, but also the world in which the individual ex- one to rate themselves on a personality inventory. ists and the events that precipitate change in the Experimental tests, such as the “go, no-go” task in individual, thus providing a and invalu- which people are told to inhibit a response to a cue able source of data for understanding the person when a stop signal tone is emitted have system- as a whole. Used in combination with other meth- atic relationships to personality traits such as im- ods, it is clear that narrative data can not only add pulse control (Logan, Schachar, & Tannock, 1997) a deeper, more complex understanding of basic and related forms of , such as de- psychological phenomena, but also account for im- linquency (Mezzacappa, Kindlon, & Earls, 2001). Using Multiple Methods in Personality Psychology 335

The interesting question that is as yet untested is Block, J. (1993). Studying personality the long way. In D. whether tests like these can be aggregated into a C. Funder & R. D. Parke (Eds.), Studying lives though reliable index of individual differences in person- time: Personality and development (pp. 9-41). Washing- ton. DC: APA science volumes. ality traits, just as the elementary cognitive tasks Borman, W., Whatie, L., Pulakos, E., & Oppler, S. (1991). have been aggregated to tap into cognitive ability. Models of supervisory performance ratings. Journal Despite the examples cited earlier, it remains of , 76, 863-872. anomalous for researchers to use more than one Bornstein, R. F. (1991). Criterion validity of objective method to investigate almost any phenomena in and projective dependency tests: A meta-analytic as- personality psychology. Too much time and effort sessment of behavioral prediction. Psychological As- have gone into reifying one or another technique as sessment, 11, 48-57. the gold standard method for assessing construct Boyatzis, R. E. (1973). Affiliation motivation. In D. C. such as traits (e.g., Hofstee, 1994) or motives (Mc- McClelland & R. S. Steele (Eds.), Human motivation: A book of readings (pp. 252-275). Morristown, NJ: Gen- Clelland et al., 1989). Also, there is a tendency to eral Learning Press. approach method variance as if it is uninteresting Campbell, D. T., & Fiske, D. W. (1959). Convergent and and an expectation that it should not play a role discriminant validation by the multitrait-multimthod in the type of information gleaned from an assess- matrix. , 56, 81-105. ment (cf. Ozer, 1986). This somewhat disrespectful Campbell, J. P. (1985). Improving the selection, classifica- approach to multiple methods quite possibly de- tion, and utilization of Army enlisted personnel: Annual rives from the article that inspired this book (e.g., report, 1985 fiscal year (Tech. Rep. NO. 746). Alexan- Cronbach & Gleser, 1953), in which the construct of deria, VA: U.S. Army Research Institute for the Be- havioral and . interest is supposed to supersede the method and Carroll, J. B. (1993). Human cognitive abilities: A survey of therefore converge in a robust fashion across di- factor-analytic studies. Cambridge, England: Cam- verse techniques of assessment. In contrast, more bridge University Press. realistic appraisals of the information taken from Conley, J. J. (1984). The hierarchy of consistency: A re- multiple methods point to a more sobering con- view and model of longitudinal findings on in- clusion that the information acquired from mul- dividual differences in intelligence, personality, and tiple methods may in fact be more independent self-opinion. Personality and Individual Differences, 5, than previously expected (Fiske, 1971; Meyer et al., 11-26. 2001; Ozer, 1986). Therefore, different methods of Craik, K. H. (1993). Accentuated, revealed, and quotid- ian personalities. Psychological Inquiry, 4, 278-281. assessment provide complementary information Cronbach, L. J., & Gleser, G. C. (1953). Assessing simi- rather than perfectly overlapping information. This larity between profiles. Psychological Bulletin, 50, 456- only reinforces the point that researchers should 473. use multiple methods in personality psychology by Dawes, R. M. (1994). House of cards: Psychology and psy- default to arrive at a more complete understanding chotherapy built on . New York: The Free Press. of their research interests, whether it is traits, mo- de St. Aubin, E. (1999). Personal : The intersec- tives, abilities, or life narratives. tion of personality and religious beliefs. Journal of Personality, 67, 1105-1139. REFERENCES Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and personality. Psychologi- Ackerman, P. L., Beier, M. E., & Boyle, M. O. (2002). In- cal Review, 95, 256-273. dividual differences in working memory within ano- Emmons, R.A., & McAdams, D. P. (1991). Personal striv- mological network of cognitive and perceptual speed ings and motive dispositions: Exploring the links. abilities. Journal of : General, Personality and Bulletin, 17, 648-654. 131, 567-589. Emmons, R. A. (1986). Personal strivings: An approach Benet-Martinex, V., & Waller, N. G. (1997). Further evi- to personality and subjective well-being. Journal of dence for the cross-culture generality of the Big Seven Personality and Social Psychology, 51, 1058-1068. Factor model: Indigenous and imported Spanish per- Emmons, R. A. (1989). The personal striving approach sonality constructs. Journal of Personality, 65, 567-598. to personality. In L. A. Pervin (Ed.), Goal in References-2 Roberts et al. in Handbook of Psychological Assessment (2006)

personality and social psychology (pp. 87-126). Hills- Humphreys, L. G. (1984). General intelligence. In C. R. dale, NJ: Lawrence Erlbaum Associates. Reynolds & R. T. Brown (Eds.), Perspectives on bias in Fiske, D. W. (1971). Measuring the concepts of personality. mental testing. New York: Plenum. Chicago: Aldine-Atherton. Jensen, A. R. (1988). The . Westport, CT: Praeger. Funder, D. C. (1987). Errors and mistakes: Evaluating Kahneman, D. (1999). Objective happiness. In D. Kahne- the accuracy of social judgment. Psychological Bulle- man & E. Diener (Eds.), Well-being: The foundations of tin, 101, 75-90. hedonic psychology (pp. 3-25). New York: Russell Sage Funder, D. C. (1991). Global traits: A neo-Allportian ap- Foundation. proach to personality. Psychological Science, 2, 31-39. King, L.A. (1995). Wishes, motives, goals, and personal Funder, D. C. (1995). On the accuracy of personality memories: Relations of measures of human motiva- judgment: A realistic approach. , tion. Journal of Personality, 63, 985-1007. 102, 652-670. King, L. A., Scollon, C. K., Ramsey, C., & Williams, T. Goff, M., & Ackerman, P. L. (1992). Personality-intelli- (2000.) Stories of life transition: Subjective well-be- gence relations: Assessment of typical intellectual en- ing and ego development in parents of children with gagement. Journal of , 84, 537- Down syndrome. Journal of Research in Personality, 34, 552. 509-536. Goldberg, L. R. (1993). The structure of phenotypic per- Klinger, E. (1975). Consequences of commitment to and sonality traits. American psychologist, 48, 26-34. disengagement from incentives. Psychological Review, Green, B. F. (1978). In defense of measurement. American 82, 223-231. Psychologist, 33, 664-670. Kolar, D. W., Funder, D. C., & Colvin C. R. (1996). Com- Gustafsson, J. E. (2002). Measurement from a hierarchi- paring the accuracy of personality judgments by the cal point of view. In H. L. Braun, D. G. Jackson, & self and knowledgeable other. Journal of Personality, D. E. Wiley (Eds.), The role of constructs in psychologi- 64, 311-337. cal and educational measurements (pp. 73-95). Mahwah, Kyllonen, P. C., & Christal, R. E. (1990). Reasoning abil- NJ: Lawrence Erlbaum Associates. ity is (little more than) working memory capacity?! Helson, R. (1992). Women’s difficult times and the re- Intelligence, 14, 389-433. writing of the life story. Psychology of Women Quar- Lewis, M. (1999). On the development of personality. terly, 16, 331-347. In L. A. Pervin & O. P. Johns (Eds.), Handbook of per- Hofstee, W. K. B. (1994). Who should own the definition sonality theory and research (2nd ed., pp. 327-346). New of personality? European Journal of Personality, 8, 149- York: The . 162. Lilienfeld, S. O., Wood, J. M., & Garb, H. N. (2000). The Hogan, R. T., & Nicholson, R. A. (1988). The meaning of scientific status of projective techniques.Psychological scores. American Psychologist, 43, 621- Science in the Public Interest, 1, 27-66. 626. Little, B. R. (1983). Personal projects: A rationale and Hogan, R. T., & Roberts, B. W. (2000). A socioanalytic method for investigation. Environment and Behavior, perspective on person/environment interaction. In 15, 273-309. W. B. Walsh, K. H. Craik, & R. H. Price (Eds.), New Logan, G. D., Schachar, R. J., & Tannock, R. (1997). Im- directions in person-environment psychology (pp. 1-24). pulsivity and . Psychological Science, Mahwah, NJ: Lawrence Erlbaum Associates. 8, 60-64. Holland, J. L. (1997). Making vocational : A theory of Mayer, J. D. (1995). A framework for the classification vocational personalities and work environments (3rd ed.). of personality components. Journal of Personality, 63, Lutz, FL: Psychological Assessment Resources. 819-877. Hooker, K., & McAdams, D. P. (2003). New directions in McAdams, D. P. (1993). The stories we live by: Personal aging research: Personality reconsidered. The Journal and the making of the self. New York: William of : Psychological Sciences, 58, P296-P304. Morrow. Hooker, K. (2002). New directions for research in per- McAdams, D. P. (1996). Personality, and the sonality and aging: A comprehensive model for link- storied self: A contemporary framework for studying ing level, structure, and processes. Journal of Research persons. Psychological Inquiry, 7, 295-321. in Personality, 36, 318-334. McAdams, D. P. (1999). The psychology of life stories. Horn, J. L., & Cattell, R. B. (1966). Refinement and test Review of General Psychology, 5, 100-122. of the theory of fluid and crystallized general intelli- McAdams, D. P., Reynolds, J., Lewis, M., Patten, A., & gences. Journal of Educational Psychology, 57, 253-270. Bowman, P. ( 2001). When bad things turn good and Using Multiple Methods in Personality Psychology References-3

good things turn bad: Sequences of redemption and Roberts B. W., & Robins, R. W. (2000). Broad disposi- contamination in life narratives and their relation to tions, broad aspirations: The intersection of person- psychosocial adaption in midlife and in stu- ality traits and major life goals. Personality and Social dents. Personality and Social Psychology Bulletin, 27, Psychology Bulletin, 26, 1284-1296. 474-485. Roberts, B. W., & Wood, D. (in press). Personality devel- McClelland, D. C. (1987). Motive dispositions: The mer- opment in the context of the neo-socioanalytic model its of operant versus respondent measures. In L. of personality. In D. Mroczek & T. Little (Eds.), Hand- Wheeler (Ed.), Review of personality and social psychol- book of personality development. Mahwah, NJ: - ogy (Vol. 1, pp. 11-41). Beverly Hills, CA: Sage. rence Erlbaum Associates. McClelland, D. C., Koestner, R., & Weinberger, J. (1989). Roberts, B. W., Bogg, T., Walton, K., Chyernyshenko, O. How do self-attributed and implicit motives differ? S., & Stark, S. (2004). A lexical approach to identify- Psychological Review, 96, 690-702. ing the lower-order structure of conscientiousness. Meyer, G. J., Finn, S. E., Eyde, L. D., Day, G. G., More- Journal of Research in Personality, 38, 164-178. land, K. L., Dies, R. R., et al. (2001). Psychological Rolfhus, E., & Ackerman, P. L. (1999). Assessing indi- testing and psychological assessment. American Psy- vidual differences in knowledge: Knowledge, intelli- chologist, 56, 128-165. gence, and related traits. Journal of Psychol- Mezzacappa, E., Kindoln, D., & Earls, F. (2001). Child ogy, 91, 511-526. and performance task assessments of execu- Rorer, L. G. (1965). The great response-style myth. Psy- tive functions in boys. Journal of Child Psychology & chological Bulletin, 63, 129-156. & Allied Disciplines, 42, 1041-1048. Schmalt, H. (1999). Assessing the achievement motive Murray, H. A. (1938). Explorations in personality. New using the grid technique. Journal of Research in Person- York: Oxford University Press. ality, 33, 109-130. Ozer, D. J. (1986). Consistency in personality. Berlin, Ger- Schwarz, N., & Bells, H. (1992). Constructing reality in many: Springer Publishing Company. social judgment. In L. L. Martin & A. Tesser (Eds.), Pals, J. L. (2005). Transforming the self: A two-dimensional The construction of social judgments (pp. 217-245). process model of identity construction within narratives of Hills­dale, NJ: Lawrence Erlbaum Associates. difficult life events. Manuscript in preparation. Smith, C. P. (Ed.). (1992). Motivation and personality: Paulhus, D. L., & Harms, P. D. (2004). Measuring cogni- Handbook of thematic . Cambridge, MA: tive ability with the overclaiming technique. Intelli- Cambridge University Press. gence, 32, 297-314. Spain, J. S., Eaton, L. G., & Funder, D. C. (2000). Perspec- Paulhus, D.L., Lysy, D., & Yik, M. S M. (1998). Self-re- tive on personality: The relative accuracy of self ver- port measures of intelligence: Are they useful as sus others for the prediction of emotion and behav- proxy measures of IQ? Journal of personality, 66, 525- ior. Journal of Personality, 68, 837-867. 554. Spangler, W. D. (1992). Validity of questionnaire and Pennebaker, J. W., & Francis, M. E. (1996). Cognitive, TAT measures of need for achievement: Two meta- emotional and language processes in disclosure. Cog- analyses. Psychological Bulletin, 112, 140-154. nition and Emotion, 10, 601-626. Winter D. G., John, O. P., Stewart, A. J., Klohen E. C., & Roberts, B. W., & Caspi, A. (2003). The cumulative con- Duncan, L. E. (1998). Traits and motives: Toward an tinuity model of personality development: Striking a integration of two traditions in personality research. balance between continuity and change in personal- Psychological Review, 105, 230-250. ity traits across the life course. In R. M. Staudinger & Winter, D. G. (1991). Measuring personality at a dis- U. Linderberger (Eds. ), Understanding human devel- tance: For scoring motives in running text. In A. J. opment: Lifespan psychology in exchange with other dis- Stewart, J. M. Healy Jr., & D. J. Ozer (Eds.), Perspec- ciplines (pp. 183-214). Dorecht, the Netherlands: Klu- tives in personality: Approaches to understanding lives wer Academic Publishers. (Vol. 3, pp. 59-89). London: Jessica Kingsley Publish- Roberts, B. W., & DelVecchio, W. F. (2000). The rank- ers. order consistency of personality traits from child- Wirtz, D., Kruger, J., Napa-Scollon, C., Diener, E. (2003). hood to old age: A quantitative review of longitudi- What to do on spring break? The role of predicted, nal studies. Psychological Bulletin, 126, 3-25. on-line, and remembered experience in future . Roberts, B. W., & Pomerantz, E. M. (in press). On traits, Psychological Science, 14, 520-524. situations and their integration: A developmental perspective. Personality and Social Psychology Review.