Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 16 June 2021

Implicit–Explicit Motive Congruence and Moderating Factors

Abstract and Keywords

This chapter reviews the literature on congruence between implicit (unconscious) and explicit (conscious) motives. The conventional wisdom that implicit and explicit motives are statistically independent is shown to be incorrect. Meta-analyses of past studies indicate that, on average, implicit and explicit motives are weakly positively correlated rather than uncorrelated. The correlation becomes stronger when methodological shortcomings of past research, such as unreliability of measurement, are overcome. Nevertheless, the relation remains modest enough that the discrepancy between implicit and explicit motives carries important information about personality congruence. The relation between implicit and explicit motives has been found to vary systematically as a function of substantive moderator variables, such as self-determination, self-monitoring, and body awareness. Motive congruence is predicted distally by satisfaction of basic needs during childhood and proximally by stress among individuals who have difficulty regulating affect. Motive congruence predicts important outcomes, including volitional strength, flow, well-being, healthy eating, and relationship stability. The chapter closes with a discussion of future research directions, such as the distinction between congruence and integration constructs.

Keywords: motive congruence, incongruence, implicit motives, explicit motives, moderators, self-determination, self-monitoring, well-being, personality, motivation

(p. 187) Introduction

A motive is a predisposition to approach a particular class of incentives, such as achievement, affiliation, or power, or to avoid a particular class of threats, such as failure, rejection, or domination by others. The study of motives began with the work of Henry Murray (1938), who pioneered a multimethod approach to the assessment of human motives. Of the methods that Murray developed, two were particularly influential—self-report motive questionnaires and the Thematic Apperception Test. Motive questionnaires consist of a series of statements regarding motivation in a particular content domain, such as achievement, affiliation, or power. The participant reports the level of his or her motivation using a set of numerical response options. In the Thematic Apperception Test, the participant is shown a series of pictures and is asked to tell a story about each one. The researcher interprets the stories in terms of their motivational content. David McClelland and other researchers later developed refined versions of the Thematic Apperception Test, now called Picture Story Exercises (PSEs), in which subjective interpretation of participants’ stories is replaced by application of experimentally derived coding systems (e.g., McClelland, Atkinson, Clark, & Lowell, 1953). After decades of research in the McClelland tradition, the conventional wisdom is that scores on questionnaire and PSE measures of motives tend to be uncorrelated, even when they concern the same content domain.

Most contemporary researchers interpret the lack of correlation between questionnaire and PSE measures as a substantive fact about underlying (p. 188) constructs. Picture Story Exercises and questionnaires are thought to assess different kinds of motive constructs that are conceptually and statistically independent. Questionnaires assess explicit motives—verbally encoded values that are consciously accessible. Picture Story Exercises assess implicit motives—spontaneously expressed concerns or preferences that are not verbally encoded or directly accessible to consciousness.

An implication of this substantive interpretation of statistical independence is that individuals differ markedly in motive congruence—the extent to which an individual’s levels of implicit and explicit motives are aligned. For the sake of illustration, we have depicted a correlation of r = .00 between implicit and explicit need for achievement (nAch) in Figure 11.1, based on hypothetical data. Each data point represents the implicit nAch and explicit nAch scores for an individual. Some individuals—those whose scores fall near the line y = x—have similar levels of implicit and explicit achievement motives; that is, the extent to which they value achievement is comparable to the level of concern with achievement that they reveal spontaneously through the stories they tell. These individuals display a high level of motive congruence. Other individuals—those whose scores fall far from the line y = x—have dissimilar levels of implicit and explicit achievement motives; the extent to which they value achievement is much higher or lower than the level of concern with achievement that they reveal through their stories. These individuals display a low level of motive congruence (or, equivalently, a high level of incongruence). Motive congruence in core life domains, such as achievement, affiliation, and power, is regarded as a fundamental indicator of personality coherence and has been posited to have important implications for well-being (McClelland, Koestner, & Weinberger, 1989; Weinberger & McClelland, 1990).

Implicit–Explicit Motive Congruence and Moderating Factors

Figure 11.1 Hypothetical data in which the correlation between implicit and explicit need for achievement (nAch) is r = .00. Individuals who are more congruent have data points that fall closer to the line y = x.

The substantive interpretation of the lack of correlation between PSE and questionnaire measures provides an unflattering portrait of the human condition—it suggests that the conscious sense of self is fundamentally divorced from underlying unconscious motivations. However, an alternative possibility is that PSEs and questionnaires are uncorrelated because of limitations of one or both assessment methods. Thus, it may be the assessment methodology, rather than the human psyche, that is in disarray. Of course, these two possibilities are not mutually exclusive. Our objective in this chapter is to review the literatures on the correlation between implicit and explicit motives and on substantive and methodological factors that moderate this relation. We also review the literatures on the antecedents and consequences of motive congruence. In the following, we begin by providing a more detailed historical overview of traditional and contemporary perspectives on the relation between PSE and questionnaire motive measures.

The Relation Between Picture Story Exercises and Questionnaire Measures of Motives

In their classic book The Achievement Motive, McClelland et al. (1953) described a study in which their PSE measure of nAch was found to be uncorrelated with a three-item questionnaire concerning effort to achieve. Although significant positive correlations emerged in some subsequent studies (e.g., deCharms, Morrison, Reitman, & McClelland, 1955), many studies failed to document a significant relation between PSE and questionnaire measures of nAch. Findings were similar for PSE and questionnaire measures of need for affiliation (nAff) and need for power (nPow). Accordingly, McClelland (1987) and others concluded that these two kinds of measures are statistically independent. McClelland et al. (1989) described the lack of relation between PSE and questionnaire measures as being among the most well-established findings in psychology.1

(p. 189) Statistical Independence as a Problem of Measurement

McClelland et al. (1989) observed that most researchers up to that point had interpreted the lack of correlation between PSEs and questionnaires as evidence that one method or the other is invalid (see also Weinberger & McClelland, 1990). The assumption underlying these interpretations is that PSEs and questionnaires are alternative potential methods of assessing the same construct and therefore ought to converge if both are valid. This assumption has been most explicit among researchers employing the multitrait–multimethod matrix framework (Burwen & Campbell, 1957; Campbell & Fiske, 1959). In this approach, convergent validity depends on robust relations between measures that concern the same content (here, motive domains) but that involve different assessment methods. Campbell and colleagues have reported that PSEs tend to show poor convergence with other methods, including self-reports. Campbell and Fiske (1959) concluded—at least as interpreted by McClelland et al. (1989)—that PSEs are therefore invalid. In contrast, Raven (1988) cited questionnaires as the cause of poor convergence, which he attributed to improper design. McClelland, similarly, has questioned the validity of questionnaires for the assessment of motives, particularly in his earlier writings (e.g., deCharms et al., 1955; McClelland, 1980).

Statistical Independence as a Substantive Fact of Personality Structure

Although McClelland questioned the validity of questionnaires for the assessment of motives, his more central and long-standing explanation of the lack of correlation was that PSEs and questionnaires assess distinct and independent constructs. McClelland’s early position was that PSEs assess motives, whereas so-called motive questionnaires assess schemata (McClelland, 1951) or values (McClelland, 1980). Later, McClelland concluded that the schemata or values assessed by questionnaires satisfy his criteria for the definition of a motive—they energize, direct, and select behavior—but he argued that PSEs and questionnaires assess different kinds of motives, which he called implicit motives and self-attributed (i.e., explicit) motives, respectively (McClelland et al., 1989).

McClelland’s theoretical explanation of statistical independence appears to consist of two complementary arguments (see also Thrash, Elliot, & Schultheiss, 2007): (a) implicit and explicit motives have different developmental antecedents (e.g., they are acquired through nonverbal and verbal forms of socialization, respectively), and (b) implicit motives, which develop first, do not influence the development of explicit motives, because implicit motives are not accessible to consciousness. In short, if implicit and explicit motives have different causes and do not influence one another, there should be no correlation between them.

In support of his substantive interpretation of the lack of relation between PSE and questionnaire measures, McClelland cited data showing that they have distinct nomological nets. DeCharms et al. (1955) provided early evidence of distinct behavioral correlates, and McClelland et al.’s (1989) literature review made a compelling case for distinct nomological nets more generally (see also Biernat, 1989; McClelland & Pilon, 1983; Schultheiss, 2001). Implicit motives, according to McClelland and other theorists, develop in early childhood through preverbal, affect-based associative learning, respond to task-based or experiential incentives, predict spontaneous behavior trends, and are introspectively inaccessible; explicit motives, in contrast, are thought to develop later in childhood through verbally mediated learning, are responsive to social–extrinsic or verbal–symbolic incentives, predict deliberate choices, and are accessible in the form of consciously articulated values.

In challenging the assumption that a single construct underlies PSE and questionnaire measures, McClelland introduced a problematic assumption of his own—that PSEs and questionnaires are perfectly valid as measures of implicit and explicit motives, respectively. To be precise, McClelland certainly recognized that neither kind of measure is perfectly valid; nevertheless, he neglected this fact when he attributed the lack of correlation between PSEs and questionnaires to the existence of distinct implicit and explicit motive constructs instead of (rather than in addition to) the imperfect relation between constructs and measures. The following quotation (in which n refers to implicit need and v refers to explicit value) illustrates his treatment of (p. 190) methodological and substantive explanations as mutually exclusive alternatives:

This lack of correlation bothers a lot of people and they have used it as an argument that therefore, since the v Achievement measures are more reliable, this proves that the n Achievement is not valid. To me, it demonstrated that these measures get at different aspects of personality—n Achievement at operant trends I called motives and v Achievement at values I called schemas.

(McClelland, 1980, p. 13)

An Integrative, General Model

We present the models in Figure 11.2 as a framework for understanding researchers’ explanations of the lack of correlation between questionnaires and PSEs, as well as the assumptions that have led to these interpretations. Models representing researchers’ explanations of the lack of correlation are illustrated in the third column of Figure 11.2; models depicting the tacit assumptions that have led to these interpretations are shown in the second column; and a general model, in which these assumptions are relaxed, is shown in the first column. Readers familiar with structural equation modeling may recognize that our notation and reasoning borrow from the structural equation modeling literature.

Implicit–Explicit Motive Congruence and Moderating Factors

Figure 11.2 Models that have been used to explain the lack of correlation between implicit and explicit motives (third column); models depicting the tacit assumptions that underlie the explanatory models (second column); and a general model that is relatively free of assumptions (first column).PSE = Picture Story Exercise; Que = questionnaire.

The general model in Figure 11.2a includes two underlying constructs (in ovals): an implicit motive (e.g., implicit nAch) and the corresponding explicit motive within the same content domain (e.g., explicit nAch). The curved arrow represents the correlation between them. The underlying implicit and explicit motive constructs are posited to influence scores on PSE and questionnaire measures, respectively (rectangles). Because the measures are posited not to be perfectly reliable or valid, each is also posited to be influenced by an error term (circles). These error terms represent all influences on the measure other than the underlying construct, including extraneous constructs, method variance, and random error.

If one begins with the general model in Figure 11.2a and imposes the assumption that there is one type of underlying motive rather than two, then one arrives at the alternative-methods model shown in Figure 11.2b, in which there is no distinction between implicit and explicit motive constructs. It is understandable that researchers who relied on this tacit model attributed the lack of correlation between measures to inadequate validity of questionnaires (p. 191) (Figure 11.2d) or PSEs (Figure 11.2e). The reasoning is sound but the premise (assumption) is not, thus leading these researchers to the questionable conclusion that one method or the other is invalid.

If one begins with the general model in Figure 11.2a and instead imposes the assumption that each measure is perfectly valid, then one arrives at the model in Figure 11.2c, in which there is no distinction between constructs and measures. Relying on this tacit model, McClelland attributed the lack of correlation between measures to the independence of the underlying constructs (Figure 11.2f). Again, the reasoning is sound but the premise is not, leading McClelland and others to the questionable conclusion that the implicit and explicit motive constructs are statistically independent.

We propose that the best route forward is to make neither assumption and to allow the general model itself to guide theory and research. This model implies that the correlation between PSEs and questionnaires is a multiplicative function of three correlations corresponding to three intervening paths (Bollen, 1989): (a) the correlation between underlying implicit and explicit motive constructs, (b) the correlation between the implicit motive construct and the PSE measure, and (c) the correlation between the explicit motive construct and the questionnaire measure. Thus, a null or weak correlation between PSE and questionnaire measures may result from a null or weak relation between underlying constructs, from poor validity of one or both measures, or from some combination of these (see Figure 2g). Our general model therefore calls for research on two kinds of influences on the relation between PSE and questionnaire measures: methodological factors concerning the relation between constructs and measures and substantive factors concerning the relation between underlying constructs. This dual emphasis parallels similar historical developments in the attitude–behavior consistency and trait–behavior consistency literatures (Kraus, 1995). In the following sections, we offer an updated perspective (see also Thrash, Maruskin, & Martin, 2012) on the implicit–explicit correlation and the methodological and substantive factors that affect it.

The Time Has Come to Retire the Independence Hypothesis

For the following reasons, we call for researchers to reject the conventional wisdom of statistical independence. First, meta-analyses of past studies now provide unambiguous evidence that the independence hypothesis is inconsistent with the data. Spangler (1992) conducted a meta-analysis and found that, on average, implicit and explicit nAch were significantly positively correlated, r = .09, p < .001. In a more recent meta-analysis, Köllner and Schultheiss (2014) reported that, overall across motive domains (achievement, affiliation, power), implicit and explicit motives were correlated at r = .13 [.08, .18]. By Cohen’s (1988) conventions (.10 = a small effect, .30 = a medium effect, .50 = a large effect), the correlation of .13 indicates a small positive effect. The 95% interval of [.08, .18] indicates that the effect is extremely statistically reliable once findings are aggregated across studies.

Second, the implicit–explicit correlation is not so weak as to be negligible. Here we disagree with Köllner and Schultheiss (2014), who argued that the observed correlation of .13 is close enough to .00 that it may be regarded as supporting the independence hypothesis. A problem with the close-enough argument is that a critic would be justified in applying this same reasoning in the opposite direction—one could argue that .13 is close enough to .30 that the hypothesis of a robust, moderate effect size is supported (see also Rosenthal & Rubin, 1994). Köllner and Schultheiss bolstered their argument by pointing out that a correlation of .13 corresponds to a proportion of shared variance (i.e., r2) of only .017 or 1.7%. We caution that the shared variance metric can make effects seem less important than they actually are in a practical sense (Rosenthal & Rubin, 1979). Many published effects in the motive literature would seem unimportant if correlations were routinely converted to a squared metric to judge their importance. Decisions about the importance of an effect call for a consideration of the theoretical and historical context in which the effect is interpreted. Our remaining arguments address such issues.

Third, it is now clear that McClelland’s theoretical arguments for statistical independence were not rigorously reasoned in the first place. Consider the argument that distinct developmental antecedents (e.g., nonverbal and verbal socialization practices) lead to statistical independence. This argument presumes that these antecedents are themselves uncorrelated. After all, if nonverbal and verbal socialization practices are correlated to some degree, then implicit and explicit nAch would be correlated as well; distinct but related developmental antecedents would produce distinct but related motives. Thus, the expectation of independence is reducible to an inadvertent assumption of independence and hence lacks theoretical grounding.

(p. 192) Consider next McClelland’s argument that implicit motives do not influence explicit motives because the former are not accessible to consciousness. From a theoretical standpoint, we question the wisdom of conceptualizing implicit motives as unconscious by definition, as if some kind of impenetrable mental barrier or defense mechanism ensures that no individuals gain insight into their implicit motives. Moreover, we are aware of no affirmative evidence (not based on null effects) that individuals lack such insight, and there is some evidence to the contrary (Köllner & Schultheiss, 2014; Sherwood, 1966). Our view is that an introspective route to congruence is available but requires negotiating a series of formidable challenges: (a) introspective awareness of implicit motive arousal, (b) veridical interpretation and attribution, and (c) reconciliation of the implicit motive with other sources of explicit values (Thrash, Cassidy, Maruskin, & Elliot, 2010). Regardless of whether this process unfolds successfully in a given individual, congruence may result from processes other than introspective access. Individuals may learn about their implicit motives indirectly, such as through feedback from others (Murray, 1938). Implicit motives could also influence explicit motives through processes that do not require conscious awareness of implicit motives, such as those based on the reinforcing emotional consequences of motive congruence (Thrash et al., 2010). As with McClelland’s first argument, here, too, McClelland’s expectation of independence is reducible to an assumption of independence—that no process exists through which separate motive systems may influence one another. Related arguments for independence framed in terms of distinct cognitive processes, distinct brain process, distinct evolutionary processes, etc., are subject to the same critique.

Fourth, McClelland had reasons to emphasize independence that were more appropriate in his historical context than in ours in the early 21st century. Consider the following quotation, in which he downplayed the possibility of a positive implicit–explicit relation: “These two types of measures are essentially independent, as they ought to be on theoretical grounds, and … when occasional correlations appear between them, they are the product of a peculiar set of circumstances related to the particular group being tested” (McClelland, 1987, p. 521). McClelland had spent much of his career attempting to convince skeptics that PSEs and questionnaires assess different constructs. In this historical context, an emphasis on independence may have been necessary to make a convincing case for discriminant validity. However, now that the discriminant validity of distinct implicit and explicit motive systems is no longer in question, there is no reason to downplay evidence of a relation between them.

Finally, the argument that implicit and explicit motive constructs are positively correlated rather than uncorrelated is more than an academic quibble; these two scenarios have very different implications for theory and application. A positive implicit–explicit correlation, even a weak one, opens up new avenues for research. Imagine, hypothetically, that individuals may indeed attain some degree of motive congruence through integrative processes but that congruence may only be detected with optimal assessment methods or among those who employ optimal integrative strategies. From this perspective, the observed implicit–explicit correlation of .13 would not be expected to capture the full relation between implicit and explicit motives; rather, it would capture only the main effect, the average relation across a haphazard assortment of optimal and nonoptimal conditions. One might expect implicit and explicit motives to be more robustly related under optimal methodological or substantive conditions and unrelated under nonoptimal conditions. Thus, a weak positive association is sufficient to motivate a search for possible moderating factors. In contrast, if one adheres to the old view that implicit and explicit motives are statistically independent, then the incentive to search for moderators is undermined.2 Indeed, the recognition that a weak positive correlation pairs nicely with moderation hypotheses (e.g., Thrash & Elliot, 2002) has already catalyzed the emerging literature on factors that influence the implicit–explicit relation. We turn next to this literature.

Methodological Factors That Influence Estimates of the Implicit–Explicit Motive Correlation

Omnibus Effect of Multiple Methodological Factors

Sherwood (1966) noted that a variety of factors, such as lack of clarity about the task of introspection, defensiveness, and social desirability, may compromise the validity of questionnaire measures, at least (p. 193) as they are typically administered. Sherwood administered questionnaire measures of nAch and nAff under special conditions designed to minimize these problems. Specifically, he taught participants about the implicit nAch and nAff constructs about which they were asked to report explicitly; he sought to maximize motivation to be accurate by framing the study as an opportunity to develop self-insight; and he conducted the study in the context of a nonevaluative relationship with the experimenter. Implicit nAch and nAff were assessed under standard conditions at the beginning of the study. Also noteworthy is a methodological refinement to which Sherwood himself drew little attention—his questionnaires were designed to correspond closely in content to the implicit motive coding systems. His findings were striking. Across two studies, the correlations between implicit and explicit measures were positive and significant for both nAch (rs = .35, .42) and nAff (rs = .40, .34). These correlations are among the strongest reported to date, suggesting that implicit and explicit motives are robustly related when care is taken to avoid problems that may compromise the validity of explicit measures. Unfortunately, it is impossible to know to what extent each of Sherwood’s various methodological refinements was responsible for his findings. Next, we turn to studies that have isolated particular methodological factors.

Correspondence of Content

Ajzen and Fishbein (1977) showed that attitude–behavior consistency is attenuated when attitudes and behaviors do not correspond closely in content or specificity. Thrash et al. (2007) suggested that a similar issue may apply to congruence between implicit and explicit motives. The relation between implicit and explicit motives may have been underestimated in past research because questionnaires and PSEs generally have not been designed to correspond directly in content. For example, many measures of explicit nAch are based on Murray’s (1938) early conceptualization of nAch (e.g., A. L. Edwards, 1959), whereas McClelland’s coding system for implicit nAch was derived empirically (McClelland et al., 1953) and deviates from Murray’s conceptualization (Koestner & McClelland, 1990).

To examine the impact of correspondence of content, Thrash et al. (2007) administered a PSE measure of implicit nAch and four questionnaire measures of explicit nAch. Implicit nAch was assessed using Heckhausen’s (1963) coding system for hope for success. Three of the measures of explicit nAch were traditional measures that had not been designed to correspond in content to Heckhausen’s coding system. The fourth measure was a new questionnaire that consisted of five pairs of items that corresponded directly to categories of Heckhausen’s coding system (need for success, instrumental activity, expectation of success, praise, and positive affect). Results indicated that the traditional measures of explicit nAch were uncorrelated with implicit nAch (rs = .00 to .02), whereas the matched-content measure of explicit nAch was positively related to implicit nAch (r = .17).

Köllner and Schultheiss (2014) stated that two studies (Ramsay & Pang, 2013; Schultheiss, Yankova, Dirlikov, & Schad, 2009) failed to replicate the finding that matching the content of implicit and explicit measures increases the implicit–explicit correlation. We evaluate this claim in the following paragraphs. The Ramsay and Pang study, a small two-condition experiment (ns = 33, 41) on the effect of picture set ambiguity on PSE validity, was not designed to be a replication of Thrash et al. (2007), which had a cross-sectional design and a large sample (N = 204). The statistical power of the Ramsay and Pang study to detect the correlation of r = .17 from Thrash et al. was .15 and .18 in the two experimental conditions. Power of at least .80 is widely recommended. Because the Ramsay and Pang study was severely underpowered to detect the effect documented by Thrash et al., it cannot be interpreted as a failure of replication.

Schultheiss et al. (2009) assessed implicit nAch, nAff, and nPow using Winter’s (1994) coding system and assessed explicit nAch, nAff, and nPow with a new measure that corresponded closely in content to the implicit measure. In addition, the explicit motive questionnaire items were assessed with respect to the picture cues used in the PSE; participants were asked to look at the picture and respond to the questions as if they were a character in the picture. Schultheiss et al. reported that there was no relation between scores on these matched measures. They concluded, “Statistical independence between both construct types can also be observed when the explicit measure of motivation is made as similar as possible to the method of implicit motive assessment” (p. 78).

One of the grounds for their conclusion was the observed implicit–explicit correlations and associated null hypothesis tests. The correlation between matched measures was r = .18, p < .05 in the power domain; r = .11, ns, in the achievement domain; and r = .12, ns, in the affiliation domain. To conclude from these findings that implicit and explicit motives are statistically independent, one would have (p. 194) to overlook the fact that the null hypothesis was rejected in the power domain. One would also have to accept (rather than fail to reject) the null hypothesis in the achievement and affiliation domains. However, the logic of null hypothesis testing does not allow the null hypothesis to be accepted. This would amount to reinterpreting the obtained estimates (.11 and .12) as if they had been .00. A second grounds for their conclusion was that an omnibus test of the overlap of the three implicit measures with the three matched explicit measures was nonsignificant. However, this omnibus test confounds a test of the three relevant correlations (reported above) with a test of six other correlations for which effects would not be expected because content domains were mismatched (e.g., the correlation between implicit nPow and explicit nAff). A fair test of matched-content measures cannot involve mismatched content domains. Thus, the failure to reject the omnibus null hypothesis is not surprising because of dilution of the relevant effects.

We draw the following conclusions from the findings of Schultheiss et al. (2009). The implicit–explicit correlations of .18, .11, and .12 obtained using matched-content measures are at least as strong as the average correlations reported in the Spangler (1992) and Köllner and Schultheiss (2014) meta-analyses, and they are similar to the correlation of .17 reported by Thrash et al. (2007) for matched measures. If the Schultheiss et al. (2009) effects were to be added to the Köllner and Schultheiss (2014) meta-analysis, they would increase, not decrease, confidence that implicit and explicit motives are positively related. In addition, the average implicit–explicit correlation based on Schultheiss et al.’s (2009) unmatched measures was r = .09, which is (modestly) lower than their correlations for matched measures. Thus, this study not only corroborates existing evidence that matched measures of implicit and explicit motives are positively related but also provides modest support for the hypothesis that content match per se increases the implicit–explicit relation.

Reliability of Measurement

All measures demonstrate some degree of random measurement error, which attenuates effect sizes. If the correlation between implicit and explicit measures is not corrected for the unreliability of the measures—and generally it is not—then the correlation between underlying constructs will be underestimated. The proper means of correction depends on the theorized measurement model—that is, the relation between the measured variables and the construct of interest. Unfortunately, measurement models are rarely explicitly specified or tested in PSE research, and the measurement models that have been proposed are often unconventional (e.g., Atkinson & Birch, 1970; for other possible models, see McClelland, 1987; Thrash et al., 2010). Because this issue remains unresolved, we present the two most widely employed approaches to disattenuation. We illustrate the first approach using data from Thrash et al. (2007) and the second using data from Schultheiss et al. (2009).

One method of disattenuation is to use confirmatory factor analysis to remove the unique error variance from particular indicators of a construct, resulting in latent variables that correspond more closely to the construct of interest. Thrash et al. (2010) used this approach to reanalyze data from Thrash et al. (2007). An implicit nAch latent variable was modeled using separate nAch scores for each of 5 stories as indicators. A nonmatched-content explicit nAch latent variable was modeled using the 3 nonmatched measures as indicators. Finally, a matched-content explicit nAch latent variable was modeled using the 10 items from the matched-content questionnaire as indicators.

Disattenuation was found to modestly increase the effect size for nonmatched measures. As noted earlier, the implicit–explicit correlations for nonmatched measures ranged from r = .00 to .02, ns. Use of latent variables increased the implicit–explicit correlation for nonmatched measures to r = .07, ns. Disattenuation had a more pronounced effect for matched measures. As noted earlier, the implicit–explicit correlation for matched measures reported by Thrash et al. (2007) was r = .17. Use of latent variables increased the implicit–explicit correlation to r = .38, p < .01. Although the impact of disattenuation was itself substantial, particularly striking are the combined effects of addressing the correspondence and measurement error problems simultaneously. With nonmatched measures and without correcting for measurement error, the implicit–explicit correlations ranged from r = .00 to .02; with matched measures and with correction for measurement error, the implicit–explicit correlation was r = .38. The latter correlation is in the range reported by Sherwood (1966), who, as noted, also addressed multiple methodological problems simultaneously.

A second approach to disattenuation is to correct an observed correlation based on the reliabilities of the two variables. A standard approach based on traditional psychometric theory is to divide a correlation by the square root of the product of the internal consistencies of the two measures (p. 195) (Guilford & Fruchter, 1978). We used results reported by Schultheiss et al. (2009) to implement this technique. Schultheiss et al. (2009) did not report internal consistency values for the PSE variables, as is customary in PSE research. Internal consistency is underestimated for PSE measures, researchers have argued, because traditional psychometric models are not appropriate for the PSE. We therefore used the internal consistencies of Schultheiss et al.’s measures of explicit nPow (Cronbach’s α = 0.64), nAff (Cronbach’s α = = 0.74), and nAch (Cronbach’s α = 0.84) as estimates of the internal consistencies of the corresponding PSE measures. This approach is reasonable in that the measures were designed to be as similar as possible. Correcting the implicit–explicit correlations for unreliability using the equation described earlier increases the implicit–explicit correlations as follows: for nPow, the correlation increases from r = .18, p < .05, to r = .28, p < .0001; for nAff, the correlation increases from r = .12, ns, to r = .16, p < .05; and for nAch, the correlation increases from r = .11, ns, to r = .13, p = .07. Thus, after disattenuation, two of the effects are significantly greater than zero and one is marginally greater. These analyses provide further evidence that unreliability causes the implicit–explicit correlation to be underestimated.

Multitrait–Multimethod Analysis

As noted earlier, Campbell and Fiske (1959) concluded, based on inspection of multitrait–multimethod matrices, that PSEs and questionnaires failed tests of convergence. However, now that the discriminant validity of implicit and explicit motives is well established, the relevant question is not whether they converge strongly enough to be considered alternative indicators of the same construct (i.e., whether the correlation approaches r = 1.00), but rather whether they converge at all (i.e., whether the correlation is greater than r = .00).

Bilsky and Schwartz (2008) used multidimensional scaling to conduct a multitrait–multimethod analysis of three previously published data sets in which both PSEs and questionnaires were used to measure motivations in the achievement, affiliation, and power domains. The aim of the analysis was to derive a spatial representation of motive domain and method facets, such that more highly correlated measures are located closer together in physical space. In all three data sets, Bilsky and Schwartz found that motive domains formed pie piece–like wedges in a two-dimensional space, whereas methods were represented by concentric circles, which varied from more implicit to more explicit or vice versa as the radius increased. This structure indicates that, as may be shown with geometry, implicit and explicit measures are more strongly related when they concern the same domain. This finding contradicts the independent-constructs perspective.

In sum, empirical evidence indicates that implicit and explicit motives are more strongly related when methods are refined by improving correspondence of content, correcting for unreliability, or using contemporary modeling techniques. The effects of methodological factors tend to be individually modest but cumulatively robust. A large variety of other methodological factors remain to be investigated. We encourage researchers to look to other literatures, such as the attitude–behavior consistency, trait–behavior consistency, and implicit–explicit attitude consistency literatures, for precedents. Epstein’s (1979) research, for instance, suggests that implicit–explicit correlations will increase if test–retest reliability is enhanced by aggregating motive scores from multiple occasions. This hypothesis awaits future research.

Substantive Variables That Moderate Implicit–Explicit Motive Congruence

A weak implicit–explicit correlation is not necessarily evidence that integrative mechanisms tend to be ineffectual and therefore negligible. Another possibility is that integrative mechanisms are effectual but operative among some, rather than all, members of the population. Imagine that the true correlation between implicit and explicit nAch is r = .35. This correlation could be the net result of combining two subgroups of individuals: one in which integrative processes are operative (resulting in a correlation of, say, r = .74 within this subgroup) and another in which integrative processes are not operative (resulting in a correlation of, say, r = .00 within this subgroup). The effect of combining these two subgroups into one group is illustrated with hypothetical data in Figure 11.3. Within the past 2 decades, researchers have begun to investigate substantive moderating variables that specify the groups of individuals (e.g., individuals with high or low levels of particular traits) among whom the implicit–explicit relation is weaker or stronger. In this and other sections, we focus on studies of implicit and explicit motives per se. For reviews of studies of congruence between motives and goals, some of which preceded the motive congruence studies reviewed here (e.g., Schultheiss & Brunstein, 1999), see Brunstein (2010), Hofer and Busch (2017), and Thrash et al. (2010). (p. 196)

Implicit–Explicit Motive Congruence and Moderating Factors

Figure 11.3 Hypothetical data showing how a relatively weak correlation may be the net result of combining two subgroups, one in which the correlation is strong and one in which the correlation equals zero.


The topic of personality congruence has been of interest not only in the motive literature but also in humanistic theories, including traditional theories, such as that of Rogers (1959), and contemporary theories, such as Deci and Ryan’s self-determination theory (Deci & Ryan, 1991). Thrash and Elliot (2002) sought to integrate motive and humanistic approaches by showing that individuals who are more self-determined display greater motive congruence.

Self-determination refers to self-regulation in accord with one’s authentic or true self. Individuals differ in self-determination, such that some individuals live according to their core interests and values, whereas others live according to external or introjected controlling influences. Thrash and Elliot (2002) argued that the experience of self-determination may reflect (at least in part) the integration of explicit values with one’s preexisting and deeply rooted implicit motivational tendencies, as opposed to the internalization of explicit values arbitrarily from the environment regardless of their fit to one’s implicit motives. As expected, self-determination was found to moderate the relation between implicit and explicit nAch. Among individuals high in self-determination, implicit nAch robustly predicted explicit nAch, r = .40, p < .01; in other words, self-determined individuals tended to be congruent. Among individuals low in self-determination, implicit and explicit nAch were largely unrelated, r = –.07, ns; individuals low in self-determination tended to be either congruent or incongruent, as would be expected by chance if these individuals internalize values regardless of their fit with implicit motives.

Hofer et al. (2010) tested the generalizability of the self-determination finding across cultures using data from Cameroon, Germany, and Hong Kong. Hofer et al. (2010) reported that self-determination moderated the relation between implicit nAch and explicit achievement goals, such that implicit nAch and explicit achievement goals were positively related among individuals high, but not low, in self-determination. This moderation effect was found to be invariant across cultures.

Multiple-Moderator Approaches

Thrash et al. (2007) argued that at least three distinct processes contribute to motive congruence: access to one’s implicit motives, integration of one’s explicit motives with one’s implicit motives, and resistance to competing sources of values. Regarding access to implicit motives, Thrash et al. (2007) argued that motive congruence may be greater among individuals higher in private body consciousness; these individuals are sensitive to bodily states and therefore may perceive the effects of implicit motive arousal. Regarding integration, Thrash et al. (2007) argued that congruence may be greater among individuals higher in preference for consistency; these individuals would be particularly motivated to reconcile their explicit motives with any rudimentary knowledge of their implicit motives. Regarding resistance to competing sources of values, Thrash et al. (2007) argued that congruence may be greater among individuals lower in self-monitoring; these individuals are less likely to monitor others’ expectations and to internalize others’ values arbitrarily. As predicted, implicit nAch was found to predict explicit nAch among individuals high but not low in private body consciousness, high but not low in preference for consistency, and low but not high in self-monitoring. Extending the private body consciousness finding, Strick and Papies (2017) showed that a mindfulness exercise that included body awareness enhanced congruence between implicit motives and explicit goals.

In the Thrash et al. (2007) data, among individuals with the most advantageous profile of traits (i.e., high in private body consciousness, high in preference for consistency, and low in self-monitoring), the correlation between implicit and explicit nAch was r = .46, p < .05; among individuals with the opposite profile of traits, the correlation was r = –.30, ns. Building on this multiple-process perspective, Thrash et al. (2010) developed a general metatheoretical framework that may be useful (p. 197) in identifying additional processes through which motive congruence may emerge.

Antecedents of Motive Congruence

Next, we review studies that address essentially the same issue as those in the prior section, except that congruence is modeled differently. The studies in the prior section concerned independent variables that moderate the relation between implicit and explicit motives. For instance, self-determination was found to moderate the relation between implicit and explicit nAch. The studies in this section concern variables that predict the discrepancy between implicit and explicit motives; that is, the implicit and explicit motive variables are reduced to a single incongruence variable, which is treated as the dependent variable. For instance, if one were to test the self-determination hypothesis in this way, one would expect to find that self-determination predicts less of a discrepancy between implicit and explicit nAch. Consistent with the difference in modeling strategies, we refer to the predictor variables in this section as antecedents rather than moderators.

Need Satisfaction

Schattke, Koestner, and Kehr (2010) examined the childhood antecedents of incongruence in adults, with hypotheses grounded in self-determination theory. These authors predicted that childhood experiences that interfere with the development of self-determination—specifically, those that thwart satisfaction of the basic needs for relatedness and autonomy—would predict incongruence later in life. Based on new analyses of an archival data set (McClelland & Pilon, 1983; Sears, Maccoby, & Levin, 1957), Schattke et al. reported that experiences involving deprivation of the need for autonomy (e.g., maternal inhibition of sexuality) or of the need for relatedness (e.g., separation from the mother during the second year of life) predicted levels of incongruence 26 years later. Consistent with the self-determination findings reported earlier, these findings suggest that explicit motives become integrated with implicit motives to the extent that the socialization environment supports satisfaction of the basic needs theorized to underlie self-determination.

Referential Competence

Whereas Schattke et al. (2010) examined distal, developmental antecedents of congruence, Schultheiss, Patalakh, Rawolle, Liening, and MacInnes (2011) examined a proximal antecedent: referential competence. These authors theorized that referential competence, the ability to verbally process nonverbally presented information, promotes alignment of explicit motives and goals with implicit motives. Although results provided limited evidence of an association of referential competence with congruence between implicit motives and explicit motives, referential competence was a consistent predictor of congruence between implicit motives and explicit goals. These results suggest that explicit goals are more responsive to referential activity than are explicit motives. We theorize that referential competence facilitates the first step toward motive congruence—introspective access—and that additional integrative processes may be required for such access to enhance congruence between implicit and explicit motives per se.

Stress and Affect Regulation

Working from the perspective of personality interactions theory, Baumann, Kaschel, and Kuhl (2005) argued that stress leads to motive incongruence among state-oriented (as opposed to action-oriented) individuals. State-oriented individuals have difficulty generating positive affect in response to demand-related stressors and/or difficulty overcoming negative affect in response to threat-related stressors. In two correlational studies and an experiment in which stress was manipulated, Baumann et al. reported that state orientation interacted with stress, such that greater stress predicted a greater discrepancy between implicit and explicit nAch among state-oriented individuals but not among action-oriented individuals.

The studies in the prior section (regarding moderators) and this section (regarding antecedents) provide additional evidence that implicit and explicit motives are not statistically independent. The positive correlation between implicit and explicit motives may be viewed as the overall or average relation. The studies in the moderator section showed that the correlation varies systematically as a function of third variables. The studies in this section, similarly, indicate that one’s standing as congruent or incongruent varies systematically as a function of predictor variables. We caution that variables modeled as antecedents are not necessarily causes of congruence. Stress and mindfulness were the only antecedents that were directly manipulated in these studies.

Consequences of Motive Congruence

Although, as we have shown, implicit and explicit motives are positively related when methodological (p. 198) shortcomings of past research are overcome, the correlation remains weak enough that the discrepancy between them has important consequences. In this section, we review the literature on variables that have been posited to be consequences of motive congruence. Whereas researchers interested in moderators and antecedents generally do not distinguish the two “directional” forms of incongruence (see the upper left and lower right corners of Figure 11.1) because a given individual’s form of incongruence is presumed to be a matter of chance, researchers interested in effects of incongruence have in some cases examined the consequences of particular forms of incongruence.

Identity Status

Marcia’s (1966) theory of identity recognizes four identity statuses: moratorium, achievement, diffusion, and foreclosure. Individuals who have actively searched for an identity have the status of moratorium if they have not yet committed themselves to an identity or the status of identity achievement if they have. Individuals who have not actively searched for an identity have the status of diffusion if they have not yet committed to an identity or the status of foreclosure if they have internalized an identity from the social environment despite lack of exploration.

Regarding individuals who have committed to an identity, Hofer, Busch, Chasiotis, and Kiessling (2006) argued that identity achievement stems from discovering one’s implicit motives and adopting explicit motives consistent with them, whereas foreclosure stems from adopting others’ explicit values regardless of fit. As predicted, Hofer, Busch, et al. found that implicit and explicit nAff interacted in the prediction of identity achievement, such that explicit nAff was a more positive predictor of identity achievement among individuals higher in implicit nAff. Implicit and explicit nAff interacted in the prediction of foreclosure, such that explicit nAff was a more positive predictor of foreclosure among individuals lower in implicit nAff. No effects emerged for the statuses that do not involve commitment to an identity (i.e., moratorium, diffusion). These findings suggest that attaining a sense of identity requires discovery of one’s implicit motives and embracing them as the foundation of one’s values.

Volitional Strength

Kehr (2004) examined the relation between motive congruence and volitional regulation within a sample of managers. Kehr posited that discrepancies between implicit and explicit motives lead to psychological conflict and that resolution of this conflict requires volitional regulation. Such regulation was posited to deplete limited volitional resources. In Kehr’s study, implicit motives were assessed using the Multi-Motive Grid (Sokolowski, Schmalt, Langens, & Puca, 2000), an instrument that we classify as implicit for present purposes but that also has some properties of explicit measures. As predicted, Kehr found that discrepancies between managers’ implicit and explicit motives, averaged across content domains (achievement, affiliation, power), predicted lower levels of volitional strength 5 months later.


Flow refers to a state in which one is completely immersed in an activity, to the point of becoming unaware of anything else (Csikszentmihalyi, 1990). Rheinberg (2008) argued that the volitional regulation necessitated by motive incongruence hinders flow experiences. Consistent with this argument, Clavadetscher (2003) found that the discrepancy between implicit and explicit motives predicted less flow among workers. Schüler (2010) argued that such effects are likely to be manifest only when the situation involves motive cues; in the presence of cues, motive conflicts that are otherwise dormant are aroused and interfere with task engagement. Schüler confirmed this hypothesis in a series of three studies, including a longitudinal study in which the dependent variable represented change in flow and an experiment in which the presence of achievement incentives was experimentally manipulated.


Well-being is the outcome that has received the most attention to date. Kehr (2004) found that implicit–explicit motive discrepancies predicted lower levels of affective well-being longitudinally. Lower levels of volitional strength mediated this effect. Baumann et al. (2005) found that incongruence between implicit and explicit nAch predicted lower levels of subjective well-being and more psychosomatic complaints. Baumann et al. also showed that incongruence mediated the effect of the Stress × State orientation interaction on these outcomes. Hofer, Chasiotis, and Campos (2006) found that incongruence in the power domain, but not in the affiliation domain, predicted lower life satisfaction in three cultures. Job, Oertig, Brandstätter, and Allemand (2010) found that motive incongruence was related to higher levels of negative affect. In studies (p. 199) of managers (Kazén & Kuhl, 2011) and teachers (Wagner, Baumann, & Hank, 2016), well-being was found to be highest among those high in both implicit and explicit nPow and lowest among those who display a particular directional form of incongruence. Among managers, well-being was lowest for those high in implicit nPow and low in explicit nPow. Among teachers, well-being was lowest for those high in explicit nPow and low in implicit nPow.

Null effects of motive congruence on well-being have also been reported (McAuley, Bond, & Ng, 2004). For this empirical reason, and based on theory about the conditions under which incongruence is more or less likely to be problematic, researchers have begun to document factors that moderate the effect of motive congruence on well-being. Motive incongruence has been found to be less problematic when the incongruent motives are not aroused through motive-relevant activity (Schüler, Job, Fröhlich, & Brandstätter, 2008), when motive expression is inhibited by a dispositional trait called activity inhibition (Langens, 2007), when the individual uses emotional disclosure as a coping strategy (Langan-Fox, Sankey, & Canty, 2009; Schüler, Job, Fröhlich, & Brandstätter, 2009), and among individuals high in self-directedness or internal locus of control (Langan-Fox et al., 2009).

Unhealthy Eating

Job et al. (2010) investigated the contribution of motive incongruence to unhealthy eating behaviors. Using the Multi-Motive Grid to assess implicit motives in a sample of middle-aged women, an overall indicator of motive incongruence across motive domains was found to be positively related to multiple measures of unhealthy eating. The relation between incongruence and unhealthy eating was partially mediated by negative affect. Job et al. proposed that volitional depletion and low private body consciousness may also function as mediators. The relation of incongruence to unhealthy eating was found to be attributable specifically to incongruence in agentic domains—achievement and particularly power.

Relationship Outcomes

Hagemeyer, Neberich, Asendorpf, and Neyer (2013) proposed that motive congruence has consequences for one’s relationships. Using data from a large sample of heterosexual couples, these researchers found that congruence in relationship-focused communal motives predicted concurrent and future relationship satisfaction and future relationship stability. Hagemeyer et al. (2013) argued that the observed effects of incongruence on relationship satisfaction and stability are most likely the result of intrapersonal frustration or ambivalence caused by conflicting need systems.

Work Motivation

Thielgen, Krumm, and Hertel (2015) examined the contribution of motive congruence to work motivation using the Multi-Motive Grid to assess implicit motives. Motive incongruence in the achievement and affiliation domains was found to predict lower levels of work motivation. Within the achievement domain, the negative effect of incongruence was stronger among younger workers, who exhibited lower levels of volitional strength. Incongruence among young workers was most detrimental if achievement incentives were present.

Rawolle, Wallis, Badham, and Kehr (2016) examined the contribution of motive incongruence to job burnout. These authors argued that incongruent individuals are more likely to choose work activities that align with their explicit motives and thwart rather than support their implicit motives, resulting in lower levels of intrinsic work motivation. Lower intrinsic motivation, in turn, was theorized to predict greater job burnout. In a cross-sectional study employing the Multi-Motive Grid, Rawolle et al. (2016) found that motive incongruence predicted greater job burnout, and low intrinsic motivation mediated this effect.

In sum, motive congruence predicts a diverse set of important outcome variables related to personal identity, well-being, relationships, and work. These outcomes are generally posited to be effects of motive congruence, but causality has not been established. Demonstrating effects of motive congruence is a challenge, because motives (as aspects of personality) and the discrepancy between them are not amenable to experimental control. Thrash et al. (2010) identified several strategies for documenting causality that may be useful in future research on motive congruence.


In this chapter, we have shown to be incorrect the conventional wisdom that implicit and explicit motives are uncorrelated. Meta-analyses show that, on average across studies, implicit and explicit motives are weakly but reliably positively correlated. When care is taken to address methodological problems of past research, the relationship becomes more robust. Nevertheless, the correlation is weak enough that the discrepancy between implicit and (p. 200) explicit motives carries important information about congruence of personality. We have reviewed evidence showing that the relation between implicit and explicit motives varies systematically as a function of moderator variables. Motive congruence also has theoretically meaningful antecedents and consequences. We are excited about the rapid development of the motive congruence literature and, in the following section, identify several important questions to be addressed in future research.

Future Directions

How Should Motive Congruence Be Modeled?

Analytic strategies for modeling congruence have varied across studies. These strategies include the following: (a) testing moderation of the implicit–explicit relation (e.g., Thrash & Elliot, 2002); (b) computing difference scores based on standardized implicit and explicit motive variables (e.g., Kehr, 2004); (c) testing whether one motive moderates the effect of the other (i.e., Implicit motive × Explicit motive interactions; e.g., Hofer, Chasiotis et al., 2006); and (d) using polynomial regression with response surface analysis (e.g., Kazén & Kuhl, 2011), a sophisticated variant of the Implicit × Explicit motive interaction approach (J. R. Edwards, 2002). In most cases, implicit and explicit motives are modeled as separate variables, but in some cases they are collapsed into a single discrepancy index. Most researchers have examined nondirectional incongruence, which results from either motive being higher than the other, whereas others have examined incongruence in a particular direction. These various operationalizations of congruence, which imply subtly different conceptualizations of congruence and vary in statistical rigor, have not yet received detailed discussion. We encourage researchers to articulate the rationale for their operationalization and its suitability to the substantive issue that is being addressed. The robustness of findings across operationalizations of congruence is also in need of greater scrutiny.

What Can Motive Researchers Learn From Freud?

It is striking how similar the issues surrounding the topic of motive congruence are to the issues that interested Freud and subsequent psychoanalysts. Indeed, if one sets aside Freud’s ideas about sexuality and defenses, which, like shiny objects, attract the most attention, the findings summarized in this chapter may be viewed as vindicating Freud’s core argument about the foundation of psychological health—where id was there ego shall be (Freud, 1990, p. 100). Yet motive researchers, like empirical researchers generally, tend not to cite Freud’s work (see Weinberger & McClelland, 1990, for an exception). It is not clear whether this inattention to Freud is based on the belief that Freud’s writings have nothing to offer or a fear that citing Freud will undermine the appearance of scientific credibility. Whatever the reason, we have found that Freud’s writings contain important insights and believe that inattention to these insights undermines rather than serves scientific credibility and progress. We therefore encourage researchers to explore more fully the theoretical and historical underpinnings of their subject matter. This recommendation brings us to our next question for future research.

May Incongruence Be Integrated?

Freud (1989) argued that an individual has three healthy options after bringing unconscious material into awareness: accept it, reject it, or sublimate it. In contrast, the prevailing assumption in the motive literature seems to be that there is one healthy option: accept it—that is, embrace one’s implicit motives as the basis of one’s explicit values. We encourage researchers to entertain the possibility that rejecting or rechanneling (sublimating) one’s implicit motives may sometimes be the healthier option, particularly in the case of implicit motives (e.g., implicit nPow or implicit avoidance motives) that do not promote, or that thwart, satisfaction of fundamental human needs. Thrash et al. (2010) proposed that a self-determined, mindful decision to reject an implicit motive represents a form of integrated incongruence that may be healthier than incongruence arising through other processes (e.g., chance). Integration, which refers to unity of structure and coordination of function, may ultimately be more important than the simpler mathematical notion of congruence or discrepancy.

May Congruence Be Unintegrated?

Regarding the literature on moderators of congruence, we have argued that integrative processes may be operative for individuals at one pole of a moderating variable and not operative for individuals at the other pole. A potentially important implication is that congruent individuals are found at both poles of the moderator, but they are congruent for different reasons. For instance, one would expect most individuals high in self-determination to be congruent as a result of an integrative process, and one (p. 201) would expect roughly half of the individuals low in self-determination to be congruent as a result of chance. In the literatures on antecedents and consequences of congruence, there has been no distinction between these two sets of congruent individuals. We speculate that unintegrated congruence—congruence that arises by chance (for a broader definition of unintegrated congruence, see Thrash et al., 2010)—may be less stable and beneficial than congruence that arises through an integrative process.

Does Congruence Vary at Other Levels of Analysis?

Researchers to date have focused on congruence at the between-person level of analysis. However, congruence between implicit and explicit motives (or between the levels of implicit and explicit motivational states) is likely to vary not only across individuals but also across time (for a particular individual), across content domains (for a particular individual), across countries (for the average individual), and so on. We believe that generalizing and extending conceptualizations of congruence to new levels of analysis will be among the most fruitful avenues for future research.


Ajzen, I., & Fishbein, M. (1977). Attitude–behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888–918. doi:10.1037/0033-2909.84.5.888Find this resource:

    Atkinson, J. W., & Birch, D. (1970). The dynamics of action. New York, NY: Wiley.Find this resource:

      Baumann, N., Kaschel, R., & Kuhl, J. (2005). Striving for unwanted goals: Stress-dependent discrepancies between explicit and implicit achievement motives reduce subjective well-being and increase psychosomatic symptoms. Journal of Personality and Social Psychology, 89, 781–799. doi:10.1037/0022-3514.89.5.781Find this resource:

        Biernat, M. (1989). Motives and values to achieve: Different constructs with different effects. Journal of Personality, 57, 69–95. doi:10.1111/j.1467–6494.1989.tb00761.xFind this resource:

          Bilsky, W., & Schwartz, S. H. (2008). Measuring motivations: Integrating content and method. Personality and Individual Differences, 44, 1738–1751. doi:10.1016/j.paid.2008.02.001Find this resource:

            Bollen, K. A. (1989). Structural equations with latent variables. New York, NY: Wiley. doi:10.1002/9781118619179Find this resource:

              Brunstein, J. C. (2010). Implicit motives and explicit goals: The role of motivational congruence in emotional well-being. In O. C. Schultheiss & J. C. Brunstein (Eds.), Implicit motives (pp. 347–374). New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195335156.003.0012Find this resource:

                Burwen, L. S., & Campbell, D. T. (1957). The generality of attitudes toward authority and nonauthority figures. Journal of Abnormal and Social Psychology, 54, 24–31. doi:10.1037/h0042663Find this resource:

                  Campbell, D. T., & Fiske, D. (1959). Convergent and discriminant validation by the multitrait–multimethod matrix. Psychological Bulletin, 56, 81–105. doi:10.1037/h0046016Find this resource:

                    Clavadetscher, C. (2003). Motivation ehrenamtlicher Arbeit im Verein Mahogany Hall, Bern [Motivation for voluntary work in the association Mahogany Hall, Bern]. Bern, Switzerland: Abschlussarbeit NDS BWL/UF, Hochschule für Technik und Architektur.Find this resource:

                      Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.Find this resource:

                        Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York, NY: Harper & Row.Find this resource:

                          deCharms, R., Morrison, H. W., Reitman, W., & McClelland, D. C. (1955). Behavioral correlates of directly and indirectly measured achievement motivation. In D. C. McClelland (Ed.), Studies in motivation (pp. 414–423). New York, NY: Appleton–Century–Crofts.Find this resource:

                            Deci, E. L., & Ryan, R. M. (1991). A motivational approach to self: Integration in personality. In R. Dienstbier (Ed.), Nebraska symposium on motivation: Perspectives on motivation (Vol. 38, pp. 237–288). Lincoln: University of Nebraska Press.Find this resource:

                              Edwards, A. L. (1959). Edwards Personality Preference Schedule manual—revised. Cleveland, OH: Psychological Corporation.Find this resource:

                                Edwards, J. R. (2002). Alternatives to difference scores: Polynomial regression analysis and response surface methodology. In F. Drasgow & N. W. Schmitt (Eds.), Advances in measurement and data analysis (pp. 350–400). San Francisco, CA: Jossey–Bass.Find this resource:

                                  Epstein, S. (1979). The stability of behavior: I. On predicting most of the people much of the time. Journal of Personality and Social Psychology, 37, 1097–1126. doi:10.1037/0022-3514.37.7.1097Find this resource:

                                    Freud, S. (1989). Five lectures on psycho-analysis: The standard edition. New York, NY: Norton. (Original work published 1909).Find this resource:

                                      Freud, S. (1990). New introductory lectures on psycho-analysis (J. Strachey, Ed.). New York, NY: Norton. (Original work published 1933).Find this resource:

                                        Guilford, J. P., & Fruchter, B. (1978). Fundamental statistics in psychology and education (6th ed.). New York, NY: McGraw–Hill.Find this resource:

                                          Hagemeyer, B., Neberich, W., Asendorpf, J. B., & Neyer, F. J. (2013). (In)congruence of implicit and explicit communal motives predicts the quality and stability of couple relationships. Journal of Personality, 81, 390–402. doi:10.1111/jopy.12016Find this resource:

                                            Heckhausen, H. (1963). Hoffnung und Furcht in der Leistungsmotivation [Hope and fear components of achievement motivation]. Meisenheim am Glan, Germany: Hain.Find this resource:

                                              Hofer, J., & Busch, H. (2017). Why Citizen Kane was unhappy: Motive–goal incongruence. Social and Personality Psychology Compass, 11, 1–11. doi:10.1111/spc3.12330Find this resource:

                                                Hofer, J., Busch, H., Bond, M. H., Kärtner, J., Kiessling, F., & Law, R. (2010). Is self-determined functioning a universal prerequisite for motive–goal congruence? Examining the domain of achievement in three cultures. Journal of Personality, 78, 747–779. doi:10.1111/j.1467–6494.2010.00632.xFind this resource:

                                                  Hofer, J., Busch, H., Chasiotis, A., & Kiessling, F. (2006). Motive congruence and interpersonal identity status. Journal of Personality, 74, 511–541. doi:10.1111/j.1467–6494.2006.00383.xFind this resource:

                                                    Hofer, J., Chasiotis, A., & Campos, D. (2006). Congruence between social values and implicit motives: Effects on life satisfaction across three cultures. European Journal of Personality, 20, 305–324. doi:10.1002/per.590Find this resource:

                                                      (p. 202) Job, V., Oertig, D., Brandstätter, V., & Allemand, M. (2010). Discrepancies between implicit and explicit motivation and unhealthy eating behavior. Journal of Personality, 78, 1209–1238. doi:10.1111/j.1467–6494.2010.00648.xFind this resource:

                                                        Kazén, M., & Kuhl, J. (2011). Discrepancies between explicit and implicit power motive decreases managers’ well-being. Motivation and Emotion, 35, 317–327 doi:10.1007/s11031-011-9219-8Find this resource:

                                                          Kehr, H. M. (2004). Implicit/explicit motive discrepancies and volitional depletion among managers. Personality and Social Psychology Bulletin, 30, 315–327. doi:10.1177/0146167203256967Find this resource:

                                                            Koestner, R., & McClelland, D. C. (1990). Perspectives on competence motivation. In L. A. Pervin (Ed.), Handbook of personality: Theory and research (pp. 527–548). New York, NY: Guilford Press.Find this resource:

                                                              Köllner, M. G., & Schultheiss, O. C. (2014). Meta-analytic evidence of low convergence between implicit and explicit measures of the needs for achievement, affiliation, and power. Frontiers in Psychology, 5, 1–20. doi:10.3389/fpsyg.2014.00826Find this resource:

                                                                Kraus, S. J. (1995). Attitudes and the prediction of behavior: A metaanalysis of the empirical literature. Personality and Social Psychology Bulletin, 21, 58–75. doi:10.1177/0146167295211007Find this resource:

                                                                  Langan-Fox, J., Sankey, M. J., & Canty, J. M. (2009). Incongruence between implicit and self-attributed achievement motives and psychological well-being: The moderating role of self-directedness, self-disclosure and locus of control. Personality and Individual Differences, 47, 99–104. doi:10.1016/j.paid.2009.02.005Find this resource:

                                                                    Langens, T. A. (2007). Congruence between implicit and explicit motives and emotional well-being: The moderating role of activity inhibition. Motivation and Emotion, 31, 49–59. doi:10.1007/s11031-006-9038-5Find this resource:

                                                                      Marcia, J. E. (1966). Development and validation of ego identity status. Journal of Personality and Social Psychology, 3, 551–558. doi:10.1037/h0023281Find this resource:

                                                                        McAuley, P. C., Bond, M. H., & Ng, I. W. (2004). Antecedents of subjective well-being in working Hong Kong adults. Journal of Psychology in Chinese Societies, 5, 25–49.Find this resource:

                                                                          McClelland, D. C. (1951). Personality. New York, NY: Sloane. doi:10.1037/10790-000Find this resource:

                                                                            McClelland, D. C. (1980). Motive dispositions. The merits of operant and respondent measures. In L. Wheeler (Ed.), Review of personality and social psychology (Vol. 1, pp. 10–41). Beverly Hills, CA: Sage.Find this resource:

                                                                              McClelland, D. C. (1987). Human motivation. New York, NY: Cambridge University Press.Find this resource:

                                                                                McClelland, D. C., Atkinson, J. W., Clark, R. A., & Lowell, E. L. (1953). The achievement motive. New York, NY: Appleton–Century–Crofts. doi:10.1037/11144-000Find this resource:

                                                                                  McClelland, D. C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit motives differ? Psychological Review, 96, 690–702. doi:10.1017/CBO9780511527937.004Find this resource:

                                                                                    McClelland, D. C., & Pilon, D. A. (1983). Sources of adult motives in patterns of parent behavior in early childhood. Journal of Personality and Social Psychology, 44, 564–574. doi:10.1037/0022-3514.44.3.564Find this resource:

                                                                                      Murray, H. A. (1938). Explorations in personality. New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195305067.001.0001Find this resource:

                                                                                        Ramsay, J. E., & Pang, J. S. (2013). Set ambiguity: A key determinant of reliability and validity in the picture story exercise. Motivation and Emotion, 37, 661–674. doi:10.1007/s11031-012-9339-9Find this resource:

                                                                                          Raven, J. (1988). Toward measures of high-level competencies: A re-examination of McClelland’s distinction between needs and values. Human Relations, 41, 281–294. doi:10.1177/001872678804100401Find this resource:

                                                                                            Rawolle, M., Wallis, M. S., Badham, R., & Kehr, H. M. (2016). No fit, no fun: The effect of motive incongruence on job burnout and the mediating role of intrinsic motivation. Personality and Individual Differences, 89, 65–68. doi:10.1016/j.paid.2015.09.030Find this resource:

                                                                                              Rheinberg, F. (2008). Intrinsic motivation and flow experience. In H. Heckhausen & J. Heckhausen (Eds.), Motivation and action (pp. 323–348). Cambridge, UK: Cambridge University Press. doi:10.1017/CBO9780511499821.014Find this resource:

                                                                                                Rogers, C. R. (1959). A theory of therapy, personality, and interpersonal relationships, as developed in the client-centered framework. In S. Koch (Ed.), Psychology: A study of a science: Vol. 3. Formulations of the person and the social context (pp. 184–215). New York, NY: McGraw–Hill.Find this resource:

                                                                                                  Rosenthal, R., & Rubin, D. B. (1979). A note on percent variance explained as a measure of importance of effects. Journal of Applied Social Psychology, 9, 395–396. doi:10.1111/j.1559–1816.1979.tb02713.xFind this resource:

                                                                                                    Rosenthal, R., & Rubin, D. B. (1994). The counternull value of an effect size: A new statistic. Psychological Science, 5, 329–334. doi:10.1111/j.1467–9280.1994.tb00281.xFind this resource:

                                                                                                      Schattke, K., Koestner, R., & Kehr, H. M. (2010). Childhood correlates of adult levels of incongruence between implicit and explicit motives. Motivation and Emotion, 35, 306–316. doi:10.1007/s11031-010-9182-9Find this resource:

                                                                                                        Schüler, J. (2010). Achievement incentives determine the effects of achievement–motive incongruence on flow experience. Motivation and Emotion, 34, 2–14. doi:10.1007/s11031-009-9150-4Find this resource:

                                                                                                          Schüler, J., Job, V., Fröhlich, S. M., & Brandstätter, V. (2008). A high implicit affiliation motive does not always make you happy: A corresponding explicit motive and corresponding behavior are further needed. Motivation and Emotion, 32, 231–242. doi:10.1007/s11031-008-9096-yFind this resource:

                                                                                                            Schüler, J., Job, V., Fröhlich, S. M., & Brandstätter, V. (2009). Dealing with a “hidden stressor”: Emotional disclosure as a coping strategy to overcome the negative effects of motive incongruence on health. Stress and Health, 25, 221–233. doi:10.1002/smi.1241Find this resource:

                                                                                                              Schultheiss, O. C. (2001). An information processing account of implicit motive arousal. In M. L. Maehr & P. Pintrich (Eds.), Advances in motivation and achievement (Vol. 12, pp. 1–41). Greenwich, CT: JAI Press.Find this resource:

                                                                                                                Schultheiss, O. C., & Brunstein, J. C. (1999). Goal imagery: Bridging the gap between implicit motives and explicit goals. Journal of Personality, 67, 1–38. doi:10.1111/1467-6494.00046Find this resource:

                                                                                                                  Schultheiss, O. C., Patalakh, M., Rawolle, M., Liening, S., & MacInnes, J. J. (2011). Referential competence is associated with motivational congruence. Journal of Research in Personality, 45, 59–70. doi:10.1016/j.jrp.2010.11.014Find this resource:

                                                                                                                    Schultheiss, O. C., Yankova, D., Dirlikov, B., & Schad, D. J. (2009). Are implicit and explicit motive measures statistically independent? A fair and balanced test using the Picture Story Exercise and a cue- and response-matched questionnaire measure. Journal of Personality Assessment, 91, 72–81. doi:10.1080/00223890802484456Find this resource:

                                                                                                                      (p. 203) Sears, R. R., Maccoby, E. E., & Levin, H. (1957). Patterns of child rearing. Evanston, IL: Row, Peterson.Find this resource:

                                                                                                                        Sherwood, J. J. (1966). Self-report and projective measures of achievement and affiliation. Journal of Consulting Psychology, 30, 329–337. doi:10.1037/h0023574Find this resource:

                                                                                                                          Sokolowski, K., Schmalt, H-D., Langens, T. A., & Puca, R. M. (2000). Assessing achievement, affiliation, and power motives all at once: The Multi-Motive Grid (MMG). Journal of Personality Assessment, 74, 126–145. doi:10.1207/S15327752JPA740109Find this resource:

                                                                                                                            Spangler, W. D. (1992). Validity of questionnaire and TAT measures of need for achievement: Two meta-analyses. Psychological Bulletin, 112, 140–154. doi:10.1037/0033-2909.112.1.140Find this resource:

                                                                                                                              Strick, M., & Papies, E. K. (2017). A brief mindfulness exercise promotes the correspondence between the implicit affiliation motive and goal setting. Personality and Social Psychology Bulletin, 43, 623–637. doi:10.1177/0146167217693611Find this resource:

                                                                                                                                Thielgen, M. M., Krumm, S., & Hertel, G. (2015). When being old pays off: Age mitigates adverse effects of low implicit–explicit motive congruency on work motivation. Journal of Career Assessment, 23, 459–480. doi:10.1177/1069072714547613Find this resource:

                                                                                                                                  Thrash, T. M., Cassidy, S. E., Maruskin, L. A., & Elliot, A. J. (2010). Factors that influence the relation between implicit and explicit motives: A general implicit–explicit congruence framework. In O. C. Schultheiss & J. C. Brunstein (Eds.), Implicit motives (pp. 308–346). New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195335156.003.0011Find this resource:

                                                                                                                                    Thrash, T. M., & Elliot, A. J. (2002). Implicit and self-attributed achievement motives: Concordance and predictive validity. Journal of Personality, 70, 729–755. doi:10.1111/1467-6494.05022Find this resource:

                                                                                                                                      Thrash, T. M., Elliot, A. J., & Schultheiss, O. C. (2007). Methodological and dispositional predictors of congruence between implicit and explicit need for achievement. Personality and Social Psychology Bulletin, 33, 961–974. doi:10.1177/0146167207301018Find this resource:

                                                                                                                                        Thrash, T. M., Maruskin, L. A., & Martin, C. C. (2012). Implicit–explicit motive congruence. In R. M. Ryan (Ed.), The Oxford handbook of human motivation (pp. 141–156). New York, NY: Oxford University Press. doi: this resource:

                                                                                                                                          Wagner, L., Baumann, N., & Hank, P. (2016). Enjoying influence on others: Congruently high implicit and explicit power motives are related to teachers’ well-being. Motivation and Emotion, 40, 69–81. doi:10.1007/s11031-015-9516-8Find this resource:

                                                                                                                                            Weinberger, J., & McClelland, D. C. (1990). Cognitive versus traditional motivational models: Irreconcilable or complementary? In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook of motivation and cognition (Vol. 2, pp. 562–597). New York, NY: Guilford Press.Find this resource:

                                                                                                                                              Winter, D. G. (1994). Manual for scoring motive imagery in running text (4th ed.). Unpublished manuscript, Department of Psychology, University of Michigan, Ann Arbor.Find this resource:


                                                                                                                                                (1) McClelland and others who have argued that implicit and explicit motives are statistically independent have sometimes qualified this claim by stating that implicit and explicit motives are “generally,” “essentially,” or “largely” independent. We acknowledge this fact but emphasize McClelland’s focus on independence for the following reasons: (a) McClelland appeared to use these qualifications for his language to be consistent with the empirical facts, but these empirical facts appear not to have influenced his theorizing; (b) McClelland often did not qualify the claim of statistical independence; (c) McClelland (1987) explicitly denied the meaningfulness of implicit–explicit correlations when they emerged; and (d) for theory to progress, it is necessary to sharpen distinctions (e.g., between “independent” and “essentially independent”) that have been obscured in the past.

                                                                                                                                                (2) Technically, moderation could be present even if the average correlation is 0.00, but in this case moderation would require a negative conditional effect at one pole of the moderator and a positive conditional effect at the other. Such a pattern may not seem sufficiently plausible to motivate a search for moderators.