Advances in Personality Science (III)
I’ve completed the book. The last part of the book wasn’t that bad, though I remain unconvinced of some of the findings in these chapters because of methodological issues I have with the way they do things (Here’s a link relevant to one of the issues I have: “Whether individual Likert items can be considered as interval-level data, or whether they should be treated as ordered-categorical data is the subject of considerable disagreement in the literature, with strong convictions on what are the most applicable methods. This disagreement can be traced back, in many respects, to the extent to which Likert items are interpreted as being ordinal data. […] Non-parametric tests should be preferred for statistical inferences […] While some commentators consider that parametric analysis is justified for a Likert scale using the Central Limit Theorem, this should be reserved for when the Likert scale has suitable symmetry and equidistance” (I’m far from sure these requirements are met). It was still interesting stuff, though I believe the chapters in the middle were the most interesting ones. I’d say that if your impression after reading some of the quotes I’ve posted is that ‘it’d be an interesting read’, you should probably read it. A few quotes from the last part of the book:
i. “until the late 1980s we had never conducted research on college students. When the NEO Personality Inventory was first published (Costa & McCrae, 1985), we heard from psychologists around the country who thought there was something wrong with our norms, because their student samples were far from average. Colleagues generously provided data from students at two West Coast universities, one East Coast, and one Southern university. A comparison of these data showed several striking effects: All the students differed substantially from our adult norms; all the subsamples showed very similar patterns; and men and women had parallel age trends (Costa & McCrae, 1989). The implication was that college students differed systematically from adults in the mean levels of many traits. […] We began our careers looking for signs of adult development and found mainly stability. Examining what might be expected to be the most volatile time of life, the teenage years, we now find—mostly stability. As Figure 9.7 shows, adolescence appears to occupy a plateau before the important changes of the next decade. But there is one extremely important difference between this plateau and that seen after age 30. As Roberts and DelVeccio (2000) show, the stability of individual differences is inversely related to age. Costa, Herbst et al. (2000) studied 40-year-olds over a 6- to 9-year interval and reported a median retest correlation of .83 across the five factors. Over the 4 years of college, Robins et al. (2001) reported a substantially lower median retest correlation of .60, and the median retest for gifted 12-yearolds was only .38 (Costa, Parker, & McCrae, 2000).
What these dramatically lower stability coefficients mean is that adolescence really is a turbulent time in which the personality traits of any given individual may change considerably. But across individuals, there is no uniform trend. Some teenagers become more agreeable—more courteous, generous, and modest—as they go through junior high and high school, but an equal number become more antagonistic, belligerent, and arrogant. Similarly, individuals’ shifts in N, E, and C appear to yield no net effect on mean levels at this age.
The exception is O, on which both boys and girls show systematic change in mean level. […] Social class certainly has marked effects on the life course, but there is little data on whether personality traits develop differently in different social groups. Physical health status in general has little effect on personality or its stability (Costa, Metter, & McCrae, 1994)” (from chapter 9)
“individuals were more compartmentalized when stress was high than when stress was low (t(13) = 2.71, p < .02). With the exception of the low vulnerability, minor events group, there was a tendency for all groups to be more compartmentalized when stress was high than when stress was low. […] among those who were experiencing high levels of stress, greater compartmentalization was associated with less negative mood.
Although these data are correlational and must be interpreted cautiously, they are consistent with the notion that increases in compartmentalization may be an effective response to stressful life events. Individuals who have the flexibility to change their type of self-organization may experience less negative mood when stressful events occur.” (from chapter 11)
“Interpersonal behavior […] involves the temporal coordination of behavior at divergent levels of analysis, from basic movements and utterances to broad action categories reflecting momentary goals and long-range plans. Even something as elemental as leaving a room, after all, requires that the room’s occupants coordinate their physical movements so as not to stumble over each other. As group action becomes more complex, the ability of group members to coordinate their activities in time becomes correspondingly more important.
To distinguish the dynamic aspects of coordination from its conventional interpretation, we employ the term synchronization. Synchronization refers to the fact that the actions, thoughts, and feelings of one person are temporally related to the actions, thoughts, and feelings of one or more other people. […] In its most basic form, synchronization refers to the coupling of behavior patterns […] Synchronization […] is likely to become more difficult as the action in question becomes more complex. It may be impossible, for example, for two unacquainted people to synchronize their efforts sufficiently to assemble a mechanical device or create a piece of art. The ability to synchronize in more complex modes requires at least some semblance of concordance in the requisite internal states of each person. […] The importance of similarity in facilitating synchronization is apparent with respect to stable characteristics such as attitudes, values, talents, temperament, and personality traits. Indeed, similarity with respect to such characteristics has been shown consistently to be among the strongest preconditions for interpersonal attraction (cf. Byrne, Clore, & Smeaton, 1986; Newcomb, 1961). By the same token, individuals avoid forming relationships with people who appear to be different from them in their personal characteristics (e.g., Rosenbaum, 1986). […]
Even if someone’s internal state is readily detectable, it may prove difficult to modify one’s own state to match it. It is hard to change one’s cognitive style or temperament, for example, regardless of how pragmatic it would be do so in preparing for an interaction with someone whose way of thinking and tempo of expression is markedly different from one’s own. There is evidence, for example, that differences in temperament can hinder effective emotional and behavioral coordination (e.g., Dunn & Plomin, 1990). In this sense, personality sets constraints on social interaction. People’s stable characteristics—traits, values, and the like—bias the choice of interaction partners and dictate the likely success of establishing relationships with those who are chosen.
But one can look at the process in reverse to ask how social interactions shape personality. Personality, after all, comes from somewhere. […] We propose that individual differences are shaped by the history of social interactions. […] In essence, the model envisions social interaction as a vehicle for coupling the dynamics of individuals. Each individual brings his or her personal dynamic tendencies to social interaction and attempts to synchronize these tendencies with his or her interaction partner. As a result of these attempts, social interaction revises the settings for each individual, or engraves entirely new settings, which then provide the foundation for subsequent social interactions. In principle, this reciprocal relation between settings of internal parameters and social interaction iterates continuously throughout social life. In reality, the engravings of some tendencies are likely to become particularly stable and thus resistant to modification in the ordinary course of social encounters. […]
With respect to modeling human dynamics, the dynamical variable (x) can be interpreted as behavior. Changes in x thus reflect variation in the intensity of behavior. The control parameter, r, corresponds to internal states (e.g., personality traits, moods, values, etc.) that shape the person’s pattern of behavior (i.e., changes in x over time). […] (α) corresponds to the strength of coupling and reflects the mutual interdependency of the relationship. When the fraction is 0, there is no coupling on the behavior level. When the fraction is 1, each person’s behavior is determined equally by his or her preceding behavior and the preceding behavior of the other person. […] The main results […of their simulations – US] were straightforward. In general, the degree of synchronization between partners’ behaviors increased both with α and similarity in r. This implies that similarity in internal states and interdependence can compensate for one another in achieving or maintaining a given level of synchronization. […] Modeling the direct synchronization of control parameters is relatively straightforward. One need only assume that on each simulation step, the value of each person’s control parameter drifts somewhat in the direction of the value of the partner’s control parameter. The rate of this drift and the size of the initial discrepancy between the values of the respective control parameters determine how quickly the control parameters begin to match. This mechanism assumes that both interaction partners can directly estimate the settings of one another’s control parameters. In many types of relationships, considerable effort may be focused on communicating or inferring these settings (cf. Jones & Davis, 1965; Kunda, 1999; Nisbett & Ross, 1980; Wegner & Vallacher, 1977). Even with such effort, however, the exact values of the relevant control parameters may be difficult or impossible to determine.
Control parameters can also become synchronized through behavioral coordination. Research concerning the facial feedback hypothesis, for instance, has established that when people are induced to mechanically adopt a specific facial configuration linked to a particular mood (e.g., disgust), they tend also to adopt the corresponding affective state (e.g., Strack, Martin, & Stepper, 1988). This matching of internal states to overt behavior is enhanced when the behavior is interpersonal in nature. Even role playing, in which a person simply follows a behavioral script in social interaction, often produces pronounced changes in attitudes and values on the part of the role player (e.g., Zimbardo, 1970). […]
Figure 12.1 shows the time course of synchronization as two maps progressively match each other’s control parameters […] This simulation was run for relatively weak coupling (α = 0.25). The x-axis corresponds to time in simulation steps, and the y-axis portrays the value of the difference between the two maps. The thin line corresponds to the difference in the dynamic variables, whereas the thicker line corresponds to the difference in r. Over time, the difference in the respective control parameters of the two maps decreases and the maps become perfectly synchronized in their behavior. This suggests that attempting behavioral synchronization with weak levels of influence and control over one another’s behavior will facilitate matching of one another’s internal states.
Figure 12.2 shows the results when the simulation was run with a stronger value of coupling (α = 0.7). Note that although coordination in behavior develops almost immediately, the control parameters fail to converge, even after 1,000 simulation steps. This is because strong coupling causes full synchronization of behavior, even for maps with quite different control parameters. Once the behavior is in full synchrony, the two maps do not have a clue that their control parameters are different. Hence, if the coupling were removed, the dynamics of the two respective maps would immediately diverge. This result suggests that using very strong influence to obtain coordination of behavior may effectively hinder synchronization at a deeper level. More generally, there is optimal level of influence and control over behavior in relationships. If influence is too weak, synchronization may fail to develop. Very strong influence, on the other hand, can prevent the development of a relationship based on mutual understanding and empathy. Although highly controlled partners may fully synchronize their behavior, they are unlikely to internalize the values of control parameters necessary to maintain such behavior in the absence of interpersonal influence. For such internalization to occur, intermediate levels of mutual influence would seem to be most effective.” (from the last chapter of the book)
No comments yet.