Normative and Distinctive Accuracy in Situation Perceptions

Abstract

To what extent do people achieve accuracy in judging others’ situations? Based on interpersonal perception models, we propose that ex situ raters may attain accuracy by judging the psychological characteristics of a situation that in situ raters have experienced according to a normative and distinctive characteristics profile. Biesanz’ social accuracy model (SAM) provides a flexible crossed-effects random coefficient modeling framework that can be applied to situation perception data. By targeting characteristics profiles with the analytical unit of the ex situ rater-situation dyad, the extent of and variation in normative and distinctive accuracy of ex situ raters can be estimated and explained by personality correlates to quantify “the good ex situ rater.” We demonstrate an SAM approach to situational accuracy with real in situ and ex situ data (402 ex situ raters judged 10 situations on 8 characteristics) and sketch future research.

Keywords

situations situation perception accuracy normative and distinctive accuracy social accuracy model

People form impressions of situations as if they were real, coherent entities (Cantor, Mischel, & Schwartz, 1982; Edwards & Templeton, 2005; Forgas, 1976; Magnusson, 1981; Rauthmann, 2012; Rauthmann, Sherman, & Funder, 2015a, 2015b; Serfass & Sherman, 2013; Sherman, Nave, & Funder, 2012, 2013). Thus, situation perception may follow principles similar to person perception (Nystedt, 1972a, 1972b, 1981; Rauthmann, 2012; Rauthmann et al., 2014). Attending to situation perception is important because knowing others’ situations can help understand and predict their behavior better. Thus, a crucial question is how good people actually are at deciphering others’ situations. Further, most research interested in the situation as the unit of analysis or as a moderating variable will need explicit measurements of situations (Rauthmann et al., 2015a), which will most often be ratings of the psychological characteristics of target situations from different sources (e.g., in situ raters: people in and affected by the situation; ex situ raters: neither present nor affected) and thus essentially tap perceptions. This opens up the door to investigate the agreement between different kinds of situation raters (Rauthmann & Sherman, 2017).

This work serves to deepen our understanding of situational accuracy as the extent to which ex situ raters can accurately judge the characteristics of situations from in situ raters with only minimal information available. Specifically, we aim to address three questions within a flexible multilevel modeling (MLM) approach:

How accurately can people generally judge others’ situations?

How strong is normative and distinctive situational accuracy?

Which broad personality traits are associated with individual differences in being normatively and distinctively accurate?

Situation Perception

Person and Situation Perception

Concepts and methods of interpersonal perception and person(ality) judgment literature can also be applied to situation perception (Nystedt, 1981; Rauthmann, 2012; Rauthmann & Sherman, 2017; Rauthmann et al., 2015a). Nonetheless, situation perception is in key respects different from person perception. First, perceptions are only unidirectional (i.e., no reciprocity as in interpersonal perceptions; Kenny, 1994). Second, situation perceptions could fluctuate more as situations are ever-changing, dynamic, and fleeting (in contrast to persons as lasting, physical beings). Third, situations cannot rate themselves, so raters are required to assess their psychological characteristics. This latter point invites the question of reality: To what extent is a situation a “real thing?” To cope with this question, Rauthmann, Sherman, and Funder (2015a) proposed three principles of psychological situation research: A psychologically relevant situation (1) only “exists” if at least one person processes it, (2) is grounded in three types of reality (physical, social, and personal), and (3) should be measured from different perspectives (i.e., in situ and ex situ raters). These principles require a better understanding of how in situ and ex situ raters agree in their perceptions of the same situation.

Situation Characteristics

People pervasively process environmental cues and form psychological situation representations imbued with meaning and interpretations (Argyle, Furnham, & Graham, 1981; Block & Block, 1981; Magnusson, 1981; Rauthmann et al., 2015a, 2015b). Psychological situations can be described with situation characteristics similarly to how persons can be described with traits (de Raad, 2004; Edwards & Templeton, 2005; Rauthmann et al., 2014). Rauthmann et al. (2015a, 2015b) argued that situation research should proceed in a variable-oriented way by using continuous dimensions of characteristics. For example, situations are assessed by asking participants to rate the extent that a certain characteristic applied (e.g., work has to be done) rather than just list cues (e.g., books lying on the desk) or classify the situation (e.g., work situation). Thus, we obtain data on what situations mean to people. This work concerns to what extent people agree in their assessments of situations’ characteristics.

Recently, Rauthmann et al. (2014) proposed to taxonomize situation characteristics into eight major domains, the Situational Eight DIAMONDS (Duty: Does work need to be done? Intellect: Is deep thinking required? Adversity: Is someone threatened? Mating: Is the situation sexually/romantically charged? pOsitivity: Is the situation enjoyable? Negativity: Could negative feelings ensue? Deception: Is mistrust an issue? Sociality: Can meaningful social interaction and relationships develop?). These eight domains, integrating most previously identified characteristic dimensions into a common framework and language, have been shown to be useful in understanding how people’s everyday situations look like (Brown & Rauthmann, 2016; Serfass & Sherman, 2015) and how personality, situations, and behavior work together (Rauthmann, 2016; Rauthmann, Sherman, Nave, & Funder, 2015c; Rauthmann et al., 2016; Rauthmann & Sherman, 2016a; Sherman, Rauthmann, Brown, Serfass, & Jones, 2015). Thus, we deem the Situational Eight a good starting point to examine situational accuracy.

Accuracy in Judging Situation Characteristics

When judging others’ situations, there can be two accuracy criteria: capturing what the typical situation is like (= normative profile) and what makes the judged situation unique (= distinctive profile as the deviation from the norm). People may be accurate with regard to none, both, or only one of these. This means we should distinguish between normative and distinctive accuracy (Biesanz, 2010; Furr, 2008), although achieving any form of situational accuracy is important in daily life. To understand others, predict their behavior, and coordinate own behavior with others, it is paramount to accurately judge others’ situations—both in terms of what situations are generally like (normative accuracy) and what makes a specific situation stand out from the average situation (distinctive accuracy). Being normatively accurate may aid navigating typical or recurrent social situations, while being distinctively accurate could aid understanding specific situations. Having distinctive situation knowledge (Sherman et al., 2012; Serfass & Sherman, 2013) may, in turn, translate into better perspective-taking skills and more empathy or at least underlie them. Thus, understanding situational accuracy is an important endeavor.

To date, however, there is only little direct research on situational accuracy. First, Rauthmann et al. (2014) had ex situ raters read brief vignettes describing the situations of in situ raters (e.g., “Going shopping with my boyfriend”), while both in situ and ex situ raters rated the situations on the DIAMONDS. The average agreement hovered around r = .50, which is quite sizable given that ex situ raters had only limited written information on in situ raters’ situations available. However, this study cannot tell us whether ex situ raters were normatively and/or distinctively accurate.

Second, Rauthmann and Sherman (2017) outlined how situational accuracy could be studied using different variance decompositions (Biesanz, 2010; Cronbach, 1955; Jussim, 2005; Kenny, 1994; Kenny, West, Malloy, & Albright, 2006). Despite their appeal, the showcased decomposition techniques cannot clearly distinguish between normative versus distinctive accuracy (Biesanz, 2010) and extract individual differences therein. For example, most techniques require piecemeal procedures to examine personality correlates of situational accuracy (i.e., effect scores need to be extracted first, which are then correlated with personality variables). This work employs an MLM approach by Biesanz (2010) to disentangle normative from distinctive situational accuracy and flexibly incorporate personality moderators.

A Social Accuracy Model (SAM) Approach to Situation Perceptions

Common analysis of variance-inspired decomposition techniques usually require multistep approaches (e.g., extracting effects, correlating them, etc.). As Biesanz (2010, p. 858) noted, such a “two-stage modeling approach is inelegant, inefficient, requires a complete and balanced design, and is restrictive in the questions that it allows to be asked,” while “through the use of [MLM], the entire analysis can be placed within a single model that will allow a richer set of substantively important questions to be addressed.” MLM thus offers several advantages. First, it can be used to analyze the entire data with robust estimations, handling missing values effectively. Second, ex situ raters’ personality traits may be introduced simultaneously into one model in higher order levels. Third, common decomposition techniques can be readily translated into MLM (e.g., Gelman, 2006; Hoffman & Rovine, 2007; Kenny et al., 2006). Lastly, variations in stimuli—here situations—can also be explicitly modeled by treating situations as random factors (Judd, Westfall, & Kenny, 2012).

To date, the most sophisticated MLM approach to address accuracy questions is Biesanz’ (2010) SAM. SAM is particularly elegant because it disentangles normative and distinctive accuracy in a straightforward manner. It embraces and reconciles the Cronbachian tradition, concerned with distinguishing normative from distinctive accuracy, and Kenny’s social relations model tradition, concerned with relations between perceivers and targets. Applying SAM to situation perception data, let the index p represents the pth perceiver, s the sth situation, and c the cth characteristic:

\begin{array}{l} Lower-level : \\ {ex}_{p s c} = β_{0 p s} + β_{1 p s} {in}_{s c} + β_{2 p s} μ_{c} + ∊_{p s c} . \end{array}

\begin{array}{l} Upper-level : \\ β_{0 p s} = β_{00} + β_{01} {Trait}_{p} + μ_{0 p} + μ_{0 s} + μ_{0 (p s)} \\ β_{1 p s} = β_{10} + β_{11} {Trait}_{p} + μ_{1 p} + μ_{1 s} + μ_{1 (p s)} \\ β_{2 p s} = β_{20} + β_{21} {Trait}_{p} + μ_{2 p} + μ_{2 s} + μ_{2 (p s)} . \end{array}

1.1

The variables as well as fixed and random effects in this crossed-random effects MLM are summarized in Table 1 (see also Biesanz, 2010, pp. 866–868). The basic rationale is that for each rater-situation dyad across all characteristics, ex situ ratings are predicted from the in situ criterion scores as well as the norm values of all characteristics (= typical/average situation: how situations are generally rated on all characteristics sampled). Thus, profile relationships across characteristics are focused on, and for these there are two kinds of accuracy (Equation 1). The parameter β_1ps provides an estimate of distinctive accuracy: How ex situ ratings uniquely capture in situ ratings, controlling for the normative profile. The parameter β_2ps provides an estimate of normative accuracy: How ex situ ratings capture the normative profile (controlled for in situ ratings).

Table 1.

Overview of Multilevel Variables and Parameters for Situation Perception Data.

Variable/Parameter	Meaning
Variables in the model
ex _psc	Ex situ ratings by perceiver p of situation s on characteristic c
In _sc	In situ criterion rating of situation s on characteristic c
μ _c	Norm value of characteristic c based on a(nother) large, preferably representative sample
Trait _p	Trait score of perceiver p (to be treated as a moderator of accuracy)
Lower-level estimates
β_0ps	Intercept
β_1ps	Level of distinctive accuracy (as the correspondence between perceiver p’s ratings and situation s’ ratings while controlling for the average characteristic profile by holding constant μ _c )
β_2ps	Level of normative accuracy (as the correspondence between perceiver p’s ratings and the average characteristic profile after partialing out situation s’ ratings)
∊ _psc	Error
Upper-level estimates
Fixed effects
β₀₀	Average intercept across perceivers and situations
β₁₀	Average distinctive accuracy slope across perceivers and situations
β₂₀	Average normative accuracy slope across perceivers and situations
Random effects
µ_1p	Perceiver p’s unique distinctive accuracy slope averaged across all situations
µ_1s	Situation s’ unique distinctive accuracy slope averaged across all perceivers
µ_1(ps)	Dyadic (plus residual) component as the specific distinctive accuracy of Perceiver p for Situation s after the grand mean (β₁₀), the perceiver main effect (µ_1p), and the situation main effect (µ_1s) of distinctive accuracy have been removed
µ_2p	Perceiver ps unique normative accuracy slope averaged across all situations
µ_2s	Situation s’ unique normative accuracy slope averaged across all perceivers
µ_2(ps)	Dyadic (plus residual) component as the specific normative accuracy of Perceiver p for Situation s after the grand mean (β₂₀), the perceiver main effect (µ_2p), and the situation main effect (µ_2s) of normative accuracy have been removed
Moderator effects
β₀₁	Slope of perceiver p’s trait associated with his/her ex situ ratings
β₁₁	Slope of perceiver p’s trait associated with his/her distinctive accuracy level
β₂₁	Slope of perceiver p’s trait associated with his/her normative accuracy level

Note. Perceiver = ex situ rater. See more information in Biesanz (2010).

The random effects of normative and distinctive accuracy for ex situ raters and situations, respectively, point toward differences in “the good ex situ rater” and “the good situation” (for personality judgment analogs, see Funder, 1995, 1999): Ex situ raters may attain more or less normative or distinctive accuracy, respectively (i.e., perceptive accuracy; Biesanz, 2010), and situations may be judged with more or less normative or distinctive accuracy, respectively (i.e., expressive accuracy; Biesanz, 2010). These relations are summarized in Table 2.

Table 2.

The “Good Ex Situ Rater” and the “Good Situation”.

Accuracy Types	Ex Situ Rater: Perceptive Accuracy	Situation: Expressive Accuracy
Normative accuracy	µ_2p: Individual differences in the extent to which a perceiver’s ratings of a situation’s characteristics capture the characteristics of the normative/average/general situation	µ_2s: Situational differences in the extent to which a situation’s characteristics are perceived similar to the characteristics of the normative/average/general situation
Distinctive accuracy	µ_1p: Individual differences in the extent to which a perceiver’s ratings of a situation’s characteristics capture the situation’s distinct, unique characteristics	µ_1s: Situational differences in the extent to which a situation’s distinct, unique characteristics are perceived

Note. Adapted from Biesanz (2010, Figure 4, p. 864).

Going even further (Equation 1.1), we can attempt to explain the individual differences in perceptive accuracy.¹ Specifically, we can model to what extent personality traits of ex situ raters moderate distinctive (β₁₁) and normative perceptive accuracy (β₂₁). Such analyses address the traits of who is a “good” (= more accurate) or “bad” (= less accurate) ex situ rater.

Current Work

Aims

This work serves three main aims in demonstrating how an MLM implementation can be fruitfully applied to situation perception data. In doing so, we focus on Biesanz’ SAM approach because it readily grants examining the magnitude of normative and distinctive situational accuracy. First, we aim to uncover how well ex situ raters can generally judge in situ raters’ situations (impressionistic accuracy). Second, we aim to disentangle normative from distinctive situational accuracy. Third, we seek to identify trait correlates in being normatively and distinctively accurate as, within SAM, personality moderators can be added to explain individual differences in perceptive accuracy.

Hypotheses

First, we expected both normative and distinctive accuracy to be sizable, although the former should be higher than the latter (e.g., Biesanz, 2010). Second, we expected sizable individual differences in perceptive accuracy: Ex situ raters should vary in their accuracies. Substantial (and meaningful) variance in perceptive accuracy is a prerequisite to introducing personality traits as predictors (or cross-level moderators, see Note 1). Third, we expected that broad personality traits, such as the Big Five, would explain perceiver slope variances. However, we did not form any clear a priori hypotheses on which Big Five traits would be positively or negatively associated with normative and distinctive accuracy, respectively. The lone exception was Neuroticism, a trait associated with higher vigilance toward potentially negative (or ambiguous) stimuli and negative biases (e.g., Hirsh & Inzlicht, 2008; Robinson, Ode, Moeller, & Goetz, 2007) so that situations may be processed more negatively than they are which may entail less accuracy in judging situations.

Method

Participants

The data used here were detailed in Rauthmann and Sherman (2016b, 2016c). Further, Rauthmann and Sherman (2017) performed various variance decomposition techniques on the data to tease apart different forms of accuracy. However, the current SAM analyses are novel and provide unique insights into situational accuracy and possible personality correlates. The Online Supplemental Materials contain all data and R codes to reproduce findings (see also osf.io/qgu6h).

In situ raters

In situ criterion data were gathered online from a German sample of N = 547 participants (407 women, 140 men; age: M = 28.01, SD = 10.47, range: 15–77 years). Participants were first asked to think about the situation they were in 24 hours earlier and then answer five questions in an open-text field: What was happening? Who was with you? What were you doing? Where were you? What time was it (approximately)? Next, participants rated their situation on the S8-I (as well as some other measures not of relevance here). From the total pool of 547 situations, we selected 10 target situations (based on the criterion to have different DIAMONDS profiles and reflect situations people could commonly experience) to be presented to a second sample, from which we obtained ex situ ratings. Table 3 shows the 10 criterion situations, along with their raw in situ ratings on the S8-I items.

Table 3.

Criterion Data: Target-situations and Their In Situ Ratings.

		Criterion Ratings In Situ
Situation Vignette (From In Situ Raters)		D	I	A	M	O	N	D	S
1	At home with my three children: Did the housework, homework help and packed for the kids their sports stuff for the evening	7	2	1	1	4	6	1	7
2	Sat in the library and read, researched, took notes	4	7	1	5	5	5	1	4
3	My spouse and I work together and he criticized me in front of our employee	7	5	7	1	1	7	6	7
4	Conversation with partner in bed while we cuddled	2	5	1	7	7	2	1	7
5	Watched a movie lying in bed and drank a glass of wine	1	1	1	1	7	1	1	1
6	My partner and I were arguing in the kitchen. Tried to make him realize that he hurt me	1	1	7	1	1	7	1	7
7	Had to introduce myself and present in English in a role play in front of everyone (one teacher and nine course participants) in a course room of a language center. I managed pretty well	6	4	5	2	5	6	7	5
8	Had breakfast and conversations with female roommate, male roommate + his girlfriend and former female roommate	1	1	1	1	7	1	1	7
9	Drinking coffee with a female friend and chatting	1	3	1	4	7	6	1	4
10	Meeting at a friend’s apartment with four friends (from the university) before a party; drinking cocktails together, chatting, and listening to music	2	4	1	5	7	2	1	7

Note. D = Duty: work has to be done; I = Intellect: deep thinking is required; A = Adversity: somebody is being threatened, accused, or criticized; M = Mating: potential romantic partners are present; O = pOsitivity: the situation is pleasant; N = Negativity: the situation contains negative feelings (e.g., stress, anxiety, guilt, etc.); D = Deception: somebody is being deceived; S = Sociality: social interactions are possible or required. Situations 1–8 were selected specifically because of their high ratings on one Situational Eight DIAMONDS dimension (see gray-shaded cells). Situations 9 and 10 were not selected on any specific basis except for being mundane, everyday situations. English translations of the German vignettes are given. The Situational Eight DIAMONDS dimensions were measured with the S8-I (Rauthmann & Sherman, 2015b).

Ex situ raters

Ex situ ratings came from N = 404 participants (300 women, 104 men; age: M = 25.2, SD = 5.51, range: 17–51 years) who rated each of the 10 situations in Table 3 on the same S8-I items as in the in situ data. Additionally, these participants completed a Big Five questionnaire.

Measures

Situation characteristics

Situations were rated in situ and ex situ on the S8-I with 1 item per DIAMONDS characteristic (Rauthmann & Sherman, 2016b), using a 7-point Likert-type scale (1 = not at all, 7 = totally). The items can be found in Table 3. The S8-I has favorable psychometric properties (Rauthmann & Sherman, 2016b) and is useful for answering substantive research questions (Sherman et al., 2015).

Personality traits

Ex situ raters’ Big Five traits were assessed with the 16-item Big Five Inventory (BFI-S16; Gerlitz & Schupp, 2005; Lang, 2005), using a 7-point Likert-type scale (1 = does not apply at all to me, 7 = applies totally to me). We obtained trait scores for openness (M = 5.50, SD = 0.95, α = .62), conscientiousness (M = 4.96, SD = 1.12, α = .68), extraversion (M = 4.65, SD = 1.32, α = .81), agreeableness (M = 5.21, SD = 1.01, α = .46), and neuroticism (M = 4.31, SD = 1.31, α = .74).

Results

SAM analyses were conducted on 32,320 observations consisting of 404 perceivers’ ex situ ratings of 10 situations on 8 characteristics. The ex situ ratings were made on a 7-point Likert-type scale from 1 to 7, with M = 3.58 (SD = 2.38). We within-characteristic centered the in situ criterion ratings on the eight characteristics by subtracting out the characteristics norm ratings on those characteristics (J. Biesanz, personal communication, March 18, 2014). These normative ratings stem from the total in situ sample (Rauthmann & Sherman, 2016b, 2016c) with the following means (also on a 7-point Likert-type scale): Duty = 3.60, Intellect = 3.48, Adversity = 1.58, Mating = 2.56, pOsitivity = 4.91, Negativity = 2.71, Deception = 1.47, and Sociality = 4.85. The norm values and personality scores were grand mean centered.²

In total, we estimated three models: Model 1 predicted ex situ ratings from in situ ratings (Aim 1), Model 2 predicted ex situ ratings from in situ ratings and normative values (Aim 2), and Model 3 added perceivers’ personality traits to Model 2 as moderators (Aim 3). Findings from Model 3 are summarized in Table 4 and random effects of Models 1–3 in Table 5.³

Table 4.

Multilevel Regression Estimates for Model 3.

Effects	Estimate	SE	95% CI	d
Fixed effects
Intercept (b ₀)	3.3928***	0.1614	[3.0475, 3.7052]
In situ rating (distinctive accuracy, b ₁)	0.4936***	0.0851	[0.3282, 0.6776]
Normative profile (normative accuracy, b ₂)	0.8789***	0.1340	[0.637, 1.1366]
Ex situ raters’ personality effects
Ex situ rating (b ₀₁)^a
Openness	0.0721***	0.0195	[0.0534, 0.0924]	0.43
Conscientiousness	−0.0360*	0.0165	[−0.053, −0.0185]	−0.26
Extraversion	−0.0109	0.0143	[−0.0237, 0.0036]	−0.09
Agreeableness	−0.0561**	0.0184	[−0.0751, −0.0388]	−0.36
Neuroticism	0.0350*	0.0141	[0.0215, 0.0475]	0.29
In situ rating (distinctive accuracy, b ₁₁)^b
Openness	−0.0040	0.0059	[−0.0139, 0.0064]	−0.09
Conscientiousness	0.0042	0.0050	[−0.0054, 0.0126]	0.12
Extraversion	0.0031	0.0043	[−0.0048, 0.0114]	0.10
Agreeableness	0.0004	0.0056	[−0.0104, 0.0108]	0.01
Neuroticism	0.0134**	0.0043	[0.0061, 0.0221]	0.44
Normative profile (normative accuracy, b ₂₁)^b
Openness	0.0140	0.0119	[−0.0101, 0.0363]	0.15
Conscientiousness	0.0416***	0.0101	[0.0238, 0.0628]	0.55
Extraversion	0.0252**	0.0088	[0.0099, 0.0427]	0.39
Agreeableness	0.0397***	0.0113	[0.0185, 0.0605]	0.47
Neuroticism	−0.0201*	0.0086	[−0.0352, −0.0034]	−0.31

Note. N = 402, N _obs = 32,160. Estimate = effect sizes of slopes; SE = standard error. 95% Confidence intervals (CIs) are based on 500 bootstrapped simulations. d = standardized effect size estimates, computed as $2 \cdot (b \cdot (\frac{S D_{x}}{S D_{y}}))$ , where b is the regression coefficient, SD_x is the standard deviation for the moderator variable, and SD_y is the random-effect standard deviation for the slope of interest. The multiplication by 2 yields an effect size equivalent to a Cohen’s d, as suggested by Gelman (2008).

^aThese estimates correspond to main effects. Because the intercept is the person’s average rating across all situations and all characteristics, personality predictors of individual differences in scale usage are captured. More open and neurotic ex situ raters had higher than average ratings and those more conscientious and agreeable had lower than average ratings. ^bThese estimates correspond to cross-level interaction effects.

*p < .05, **p < .01, ***p < .001, using estimate/SE as t-statistic on 384 degrees of freedom.

Table 5.

Summary of Random Effects.

	SDs
Models	Intercept	In Situ Rating	Normative Profile
Model 1
Perceivers	0.31	0.05	—
Situations	0.72	0.48	—
Residual	1.91	—	—
Model 2
Perceivers	0.33	0.08	0.19
Situations	0.51	0.27	0.42
Residual	1.54	—	—
Model 3
Perceivers	0.31	0.08	0.17
Situations	0.51	0.27	0.42
Residual	1.54	—	—

Note. N = 402, N _obs = 32,160. SD = standard deviation. See text for details on Models 1–3. Perceivers = ex situ raters.

Model 1: Impressionistic Accuracy

Model 1 examined the overall effect of accuracy (i.e., impressionistic accuracy) by predicting ex situ ratings only from the in situ data. Intercepts and slopes were allowed to vary across perceivers and across situations (i.e., different perceivers and different situations could vary in their accuracy levels). There was a statistically significant fixed effect (b = 0.52, 95% confidence interval [CI] = [0.23, 0.80], t = 3.40) indicating that, on average, ex situ raters accurately judged the pattern of characteristics of the situations.

The SDs of perceiver effects were 0.30 [0.27, 0.33] for intercepts and 0.05 [0.04, 0.07] for slopes. The SDs of situation effects were .72 [0.42, 1.06] for intercepts and 0.48 [0.27, 0.74] for slopes. The residual SD was 1.91. Overall, this indicates that there was sizable variation in accuracies between situations, while the variation in perceiver accuracies was not as large (see Figure 1). This indicates that there may be more differences in being a “good situation” (i.e., situations that are generally judged with more accuracy) than being a “good ex situ rater” (i.e., ex situ raters that generally judge situations with more accuracy).

Figure 1.

Slope density distributions for Model 1.

Model 2: Normative and Distinctive Accuracy

Model 2 added the normative characteristics profile as both a fixed- and random-effects predictor of ex situ ratings. Including the normative profile allows simultaneously estimating normative and distinctive accuracy (Biesanz, 2010; Cronbach, 1955; Furr, 2009). By estimating them as random effects, we allowed both normative and distinctive accuracy slopes to vary across ex situ raters and across situations. The results showed statistically significant effects of both normative (b = 0.88 [0.61, 1.15], t = 6.56) and distinctive (b = 0.49 [0.31, 0.68], t = 5.82) accuracy. Thus, on average, perceivers were both normatively and distinctively accurate in judging the characteristics profiles of situations.

The SDs of the perceiver effects were 0.33 [0.30, 0.36] for intercepts, 0.19 [0.17, 0.21] for normative accuracy slopes, and .08 [0.07, 0.09] for distinctive accuracy slopes. The SDs of situation effects were 0.51 [0.29, 0.75] for intercepts, 0.42 [0.24, 0.57] for normative accuracy slopes, and 0.27 [0.15, 0.41] for distinctive accuracy slopes. Thus, there was substantial variation in both normative and distinctive accuracy between situations, with smaller but nontrivial variation in normative and distinctive accuracy between ex situ raters (see Figure 2).

Figure 2.

Slope density distributions for Model 2.

Model 3: Personality Correlates

Model 3 examined Big Five traits as moderators of normative and distinctive accuracy. Centered scores of the personality variables and their interaction terms were added to Model 2 as fixed effects. The results from this analysis are summarized in Table 4 (with random effects in Table 5). As can be seen, the effects of normative accuracy were indeed associated with ex situ raters’ personality: Those higher on conscientiousness, extraversion, and agreeableness and lower on neuroticism were more normatively accurate, and those higher on neuroticism more distinctively accurate.

Discussion

This work used SAM analyses to disentangle normative from distinctive situational accuracy and additionally examine personality moderators. First, in line with our expectations, we found both substantial normative and distinctive accuracy, and the former was stronger in all three estimated models. Second, the variances in normative and distinctive perceptive accuracy were not as sizable as those of normative and distinctive expressive accuracy. Thus, on average, there were only small differences between ex situ raters in their accuracy levels (Figures 1 and 2). Because of the relatively large sample size of ex situ raters, we could nonetheless associate those small differences with the Big Five. Conscientiousness, extraversion, agreeableness, and neuroticism emerged as significant predictors, but not openness. Additionally, these traits predicted mostly only normative perceptive accuracy.

Magnitude of Accuracy

How strong was accuracy? This question can be answered by comparing the unstandardized effect estimates found here to those from the person perception literature. First, the level of normative accuracy (0.88) was similar to those found in person perception literature, while the level of distinctive accuracy (0.49) was higher (i.e., often around 0.10–0.30; e.g., Biesanz & Human, 2010; Biesanz et al., 2011; Chan, Rogers, Parisotto, & Biesanz, 2011; Human & Biesanz, 2011, 2012; Lorenzo, Biesanz, & Human, 2010). Second, because standardized effect sizes are somewhat problematic for Level 1 effects in MLM, we only report unstandardized regression coefficients for effects at this level (as is common in SAM analyses). The following is intended to aid in interpretation of these effects. The intercept is the average rating made by all raters across all situations and across all characteristics. The distinctive coefficient means that a 1-point increase in the in situ ratings yielded a 0.49 increase in the ex situ ratings (across all characteristics), controlling for all other fixed effects in the model (which includes the normative profile). In other words, the 0.49 value is an index of the average ex situ rater’s sensitivity to changes in the criterion (in situ), where 1.00 would be perfect correspondence. Again, this effect is across all situations and all characteristics. The normative coefficient means that a 1-point increase in the average situation (in situ ratings) yielded a 0.88 increase in ex situ ratings (again, across all situations and characteristics). This is close to perfect correspondence.⁴ Taken together, the accuracy levels found are actually quite high, especially considering that ex situ raters had only limited information available (Table 3).

Interindividual and Intersituational Differences

Differences between persons

We did not find strong individual differences in perceptive accuracy despite a relatively large sample that should have created sufficient variation. This may be the case because people are, on average, accurate in judging others’ situations—and there is little (but still nonnegligible) room for variation. After all, it is adaptive to perceive situations as most other people do because this enables joint communication and coordination in a socially shared reality (Miller, 2007; Rauthmann et al., 2015a). Indeed, person(ality) perception literature also finds small variation in perceptive accuracy (Biesanz, 2010; Kenny, 1994), explaining why it has been so difficult to identify “the good judge.” Small perceptive accuracy differences may thus generalize across judging persons and situations.

Nonetheless, some differences between persons still emerged. For example, interindividual differences in normative (but not distinctive) perceptive accuracy could be explained by the Big Five (except for openness), with small- to medium-effect sizes (see d in Table 4). On the other hand, only (high) neuroticism emerged as a predictor of distinctive perceptive accuracy. One interpretation of this pattern of findings could be that those with a more normative personality profile (i.e., high conscientiousness, extraversion, and agreeableness, with low neuroticism) tend to use situational normativeness to achieve accuracy. This is similar to findings in the person(ality) perception literature indicating that well-adjusted individuals tend to be accurate perceivers of what others are generally like (Human & Biesanz, 2011). In contrast, those with more nonnormative personalities (i.e., high neuroticism) may tend to be more distinctively accurate, probably because they deviate more from the normative profile. Interestingly, this improvement in distinctive accuracy for those high on neuroticism is more than negated by the losses in normative accuracy (i.e., 0.01 vs. −0.02). However, individuals high in neuroticism may also be better in distinctive cue detection due to their hypersensitivity and vigilance (Allen & Badcock, 2003). Guillaume et al. (2016) found that situations around the world were typically social and mildly pleasant. Thus, as negative situations seem less normative, distinctive accuracy could be achieved by a stronger focus on negative aspects.

Together, our data suggest that personality differences may matter most when using normative situation knowledge. This is intriguing because traditionally unique and distinctive patterns of perception have been associated with personality (Sherman et al., 2012). However, such research only concerned how people (uniquely) perceive situations and not whether those perceptions align with any criterion variables. Thus, our findings provide novel insights into situational accuracy.

Differences between situations

An interesting finding is the relatively large variance in situations’ expressive accuracy: There were wide intersituational differences in being judged normatively and distinctively accurately. This finding, again, corresponds to person(ality) perception literature which also find levels of expressive accuracy larger than perceptive accuracy (Human & Biesanz, 2013). We insert the caveat, however, that we only had a limited set of situations that were not selected to be homogeneous because we wanted meaningful variance. Nonetheless, it is striking how much more situation variance there was than rater variance. If we had sampled more situations and had other information available (e.g., physicobiological cues of situations; “style” characteristics such as base rate or situation strength; situation class membership), we could also attempt to explain intersituational differences in accuracy slopes the same way as interindividual accuracy differences were explained here by traits. Examining explanatory variables of “the good (or bad) situation” as moderators of expressive accuracy may be a fruitful direction for future research.

A Note on Statistical Modeling

We have demonstrated how SAM can be applied to situation perception data. It may not have escaped the notice of those initiated in MLM that other data-analytical choices could have been made. The accuracies uncovered reside at the ex situ rater and situation level. However, one could also quantify accuracies at other levels. First, running a model where we estimate random intercepts and slopes for characteristics and for situations (i.e., not for raters) yields accuracies at the level of situations and characteristics. In other words, for every situation separately the ex situ profile is predicted from the in situ profile. So for each situation, this regression would be based on 404 raters × 8 characteristics = 3,232 pairs of scores (across all raters and characteristics). Additionally, because we have a cross-classified MLM, for each characteristic we get a regression based on 404 raters × 10 situations = 4,040 pairs of scores across all situations and perceivers.

Second, as another alternative, we could estimate random intercepts and slopes for perceivers and characteristics, yielding accuracies at the level of raters and characteristics. Thus, for each rater, we get a regression based on 10 situations × 8 characteristics = 80 pairs of scores, and for each characteristic a regression based on 404 raters × 10 situations = 4,400 pairs of scores. Findings of these models are compiled in the Online Supplemental Materials (together with R code) for interested readers.

Although these models can make interesting sense of the data (e.g., looking at normative and distinctive characteristic accuracy), they are conceptually and statistically different from SAM (J. Biesanz, personal communication). In SAM, characteristics are fixed, while in these other analyses they are random. These different sets of analyses create the problem that we cannot include moderators of normative accuracies, though distinctive accuracies are not affected. Additionally, cross-random effects need to be orthogonal, so we are not able to model the effect of characteristics changing by situation and/or rater. Hence, SAM represents a more flexible modeling approach. For the sake of completeness and to stimulate future research on MLM implementations in situational accuracy studies, we wanted to alert to different but nonequivalent modeling procedures here.

Limitations and Future Directions

The limitations of this work point toward future research that is needed to overcome them. First, we have employed a relatively limited set of target situations and characteristics which may suggest lack of stimulus sampling (Judd et al., 2012; Wells & Windschitl, 1999; cf. Biesanz, 2010). This was mainly the case to reduce participant burden. However, given that ex situ raters’ variance in accuracy estimates was not too high, future research could use fewer ex situ raters, but have them rate more situations on more characteristics (e.g., Situational Eight on 32 items: Rauthmann et al., 2014; on 24 items: Rauthmann & Sherman, 2016c; on 89 items from the Riverside Situational Q-sort: Sherman et al., 2010; Wagerman & Funder, 2009). If more situations are available, moderators of expressive situational accuracy can be examined.

Second, we used only one normative profile for the “average situation.” However, the average DIAMONDS profile may be different for different classes of situations (e.g., van Heck’s, 1984, 1989 10 types of situations). It did not make sense to distinguish different classes within our sampled situations, but future research may seek to sample different situations within different classes (and possibly derive normative profiles for each class). It will be interesting to examine accuracy slopes for different classes and whether some people show high perceptive accuracy across all or only within specific classes.

Third, future research could sample several situations per in situ rater (we used just one per person), preferably a representative set of situations in daily life. This allows examining idiosyncrasies in in situ ratings and to what extent ex situ raters might pick up on those.

Lastly, future research could use traits other than the Big Five, such as empathy, perspective-taking ability, and socioemotional competencies to explain interindividual accuracy differences. These may be more closely tied to accurately judging distinctive situation profiles and thus increase the chance of predicting distinctive perceptive accuracy.

Conclusion

We demonstrated that person perception models, such as SAM, may be fruitfully applied to situation perception data, making it possible to examine normative and distinctive accuracies concerning ex situ raters (perceptive accuracy) and situations (expressive accuracy)— a distinction that has so far not been made in situations literature. This work thus provides a first window into how strong normative and distinctive situational accuracy are and which person variables may explain them. These findings may be important for research seeking to understand the basis of good perspective-taking skills.

Footnotes

Acknowledgments

We thank Jeremy Biesanz for valuable advice on data-analytical issues concerning SAM.

Author Note

All findings and R codes as well as additional analyses can be found openly at osf.io/qgu6h as well as in the Online Supplemental Materials.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

The supplemental material is available in the online version of the article.

Notes

References

Allen

N. B.

Badcock

P. B. T.

(2003). The social risk hypothesis of depressed mood: Evolutionary, psychosocial, and neurobiological perspectives. Psychological Bulletin, 129, 887–913.

Argyle

Furnham

Graham

J. A.

(1981). Social situations. Cambridge, MA: Cambridge University Press.

Bates

Maechler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1–48.

Biesanz

J. C.

(2010). The social accuracy model of interpersonal perception: Assessing individual differences in perceptive and expressive accuracy. Multivariate Behavioral Research, 45, 853–885.

Biesanz

J. C.

Human

L. J.

(2010). The cost of forming more accurate impressions accuracy-motivated perceivers see the personality of others more distinctively but less normatively than perceivers without an explicit goal. Psychological Science, 21, 589–594.

Biesanz

J. C.

Human

L. J.

Paquin

A. C.

Chan

Parisotto

K. L.

Sarracino

Gillis

R. L.

(2011). Do we know when our impressions of others are valid? Evidence for realistic accuracy awareness in first impressions of personality. Social Psychological and Personality Science, 2, 452–459.

Block

(1981). Studying situational dimensions: A grand perspective and some limited empiricism. In Magnusson

(Ed.), Toward a psychology of situations: An interactional perspective (pp. 85–106). Hillsdale, NJ: Lawrence Erlbaum.

Brown

N. A.

Rauthmann

J. F.

(2016). Situation characteristics are age graded: Mean-level patterns of the situational eight DIAMONDS across the life span. Social Psychological and Personality Science, 7, 667–679.

Cantor

Mischel

Schwartz

(1982). A prototype analysis of psychological situations. Cognitive Psychology, 14, 45–77.

10.

Chan

Rogers

K. H.

Parisotto

K. L.

Biesanz

J. C.

(2011). Forming first impressions: The role of gender and normative accuracy in personality perception. Journal of Research in Personality, 45, 117–120.

11.

Cronbach

L. J.

(1955). Processes affecting scores on “understanding of others” and “assumed similarity.” Psychological Bulletin, 52, 177–193.

12.

de Raad

(2004). Situations that matter to personality. In Eliasz

Hampson

S. E.

de Raad

(Eds.), Advances in personality psychology (Vol. 2, pp. 179–204). Hove, England: Routledge.

13.

Edwards

J. A.

Templeton

(2005). The structure of perceived qualities of situations. European Journal of Social Psychology, 35, 705–723.

14.

Forgas

J. P.

(1976). The perception of social episodes: Categorical and dimensional representations in two different social milieus. Journal of Personality and Social Psychology, 34, 199–209.

15.

Funder

D. C.

(1995). On the accuracy of personality judgment: A realistic approach. Psychological Review, 102, 652–670.

16.

Funder

D. C.

(1999). Personality judgment: A realistic approach to person perception. San Diego, CA: Academic Press.

17.

Furr

R M.

(2008). A framework for profile similarity: Integrating similarity, normativeness, and distinctiveness. Journal of Personality, 76, 1267–1316.

18.

Furr

R. M.

(2009). Profile analysis in person-situation integration. Journal of Research in Personality, 43, 196–207.

19.

Gelman

(2006). Analysis of variance. New Palgrave Dictionary of Economics (2nd ed.). Retrieved from http://www.stat.columbia.edu/∼gelman/research/published/econanova3.pdf

20.

Gelman

(2008). Scaling regression inputs by dividing by two standard deviations. Statistics in Medicine, 27, 2865–2873.

21.

Gerlitz

J.-Y.

Schupp

(2005). Zur Erhebung der Big-Five-basierten Persönlichkeitsmerkmale im SOEP. Research Notes. Retrieved from http://www.diw.de/documents/publikationen/73/diw_01.c.43490.de/rn4.pdf

22.

Guillaume

Baranski

Todd

Bastian

Bronin

Ivanova

… Funder

D. C.

(2016). The world at 7: Comparing the experience of situations across 20 countries. Journal of Personality, 84, 493–509.

23.

Hirsh

J. B.

Inzlicht

(2008). The devil you know: Neuroticism predicts neural response to uncertainty. Psychological Science, 19, 962–967.

24.

Hoffman

Rovine

M. J.

(2007). Multilevel models for the experimental psychologist: Foundations and illustrative examples. Behavior Research Methods, 39, 101–117.

25.

Human

L. J.

Biesanz

J. C.

(2011). Through the looking glass clearly: Accuracy and assumed similarity in well-adjusted individuals’ first impressions. Journal of Personality and Social Psychology, 100, 349–364.

26.

Human

L. J.

Biesanz

J. C.

(2012). Accuracy and assumed similarity in first impressions of personality: Differing associations at different levels of analysis. Journal of Research in Personality, 46, 106–110.

27.

Human

L. J.

Biesanz

J. C.

(2013). Targeting the good target: An integrative review of the characteristics and consequences of being accurately perceived. Personality and Social Psychology Review, 17, 248–272.

28.

Judd

C. M.

Westfall

Kenny

D. A.

(2012). Treating stimuli as a random factor in social psychology: A new and comprehensive solution to a pervasive but largely ignored problem. Journal of Personality and Social Psychology, 103, 54–69.

29.

Jussim

(2005). Accuracy: Criticisms, controversies, criteria, components, and cognitive processes. Advances in Experimental Social Psychology, 37, 1–93.

30.

Kenny

D. A.

(1994). Interpersonal perception: A social relations analysis. New York, NY: Guilford Press.

31.

Kenny

D. A.

West

T. V.

Malloy

T. E.

Albright

(2006). Componential analysis of interpersonal perception data. Personality and Social Psychology Review, 10, 282–294.

32.

Lang

F. R.

(2005). Erfassung des kognitiven Leistungspotenzials und der «Big Five» mit Computer-Assisted-Personal-Interviewing (CAPI): Zur Reliabilität und Validität zweier ultrakurzer Tests und des BFI-S. Research Notes. Retrieved from http://www.diw.de/documents/publikationen/73/diw_01.c.43950.de/rn9.pdf

33.

Lorenzo

G. L.

Biesanz

J. C.

Human

L. J.

(2010). What is beautiful is good and more accurately understood physical attractiveness and accuracy in first impressions of personality. Psychological Science, 21, 1777–1782.

34.

Magnusson

(1981). Toward a psychology of situations: An interactional perspective. Hillsdale, NJ: Lawrence Erlbaum.

35.

Miller

G. F.

(2007). Reconciling evolutionary psychology and ecological psychology: How to perceive fitness affordances. Acta Psychologica Sinica, 39, 546–555.

36.

Nystedt

(1972a). A modified lens model: A study of the interaction between an individual and the ecology. Perceptual and Motor Skills, 34, 479–498.

37.

Nystedt

(1972b). Predictive accuracy and utilization of cues: Study of the interaction between an individual’s cognitive organization and ecological structure. Perceptual and Motor Skills, 34, 171–180.

38.

Nystedt

(1981). A model for studying the interaction between the objective situation and a person’s construction of the situation. In Magnusson

(Ed.), Toward a psychology of situations: An interactional perspective (pp. 375–391). New York, NY: Academic Press.

39.

Rauthmann

J. F.

(2012). You say the party is dull, I say it is lively: A componential approach to how situations are perceived to disentangle perceiver, situation, and perceiver x situation variance. Social Psychological and Personality Science, 3, 519–528.

40.

Rauthmann

J. F.

(2016). Motivational factors in the perception of psychological situation characteristics. Social and Personality Psychology Compass, 10, 92–108.

41.

Rauthmann

J. F.

Gallardo-Pujol

Guillaume

E. M.

Todd

Nave

C. S.

Sherman

R. A.

… Funder

D. C.

(2014). The Situational Eight DIAMONDS: A taxonomy of major dimensions of situation characteristics. Journal of Personality and Social Psychology, 107, 677–718.

42.

Rauthmann

J. F.

Jones

A. B.

Sherman

R. A.

(2016). Directionality of person-situation transactions: Are there spill-overs among and between situation experiences and personality states? Personality and Social Psychology Bulletin, 42, 893–909.

43.

Rauthmann

J. F.

Sherman

R. A.

(2016a). Situation change: Stability and change of situation variables between and within persons. Frontiers in Psychology, 6, 1938.

44.

Rauthmann

J. F.

Sherman

R. A.

(2016b). Ultra-brief measures for the situational eight DIAMONDS domains. European Journal of Psychological Assessment, 32, 165–174.

45.

Rauthmann

J. F.

Sherman

R. A.

(2016c). Measuring the Situational Eight DIAMONDS characteristics of situations: An optimization of the RSQ-8 to the S8*. European Journal of Psychological Assessment, 32, 155–164.

46.

Rauthmann

J. F.

Sherman

R. A.

(2017). Towards a research agenda for the study of psychological situations: A variance componential framework for situation perceptions. Manuscript in revision.

47.

Rauthmann

J. F.

Sherman

R. A.

Funder

D. C.

(2015a). Principles of situation research: Towards a better understanding of psychological situations. Target Article. European Journal of Personality, 29, 363–381.

48.

Rauthmann

J. F.

Sherman

R. A.

Funder

D. C.

(2015b). New horizons in research on psychological situations and environments. Rejoinder to Target Article: European Journal of Personality, 29, 382–432.

49.

Rauthmann

J. F.

Sherman

R. A.

Nave

C. S.

Funder

D. C.

(2015c). Personality-driven situation experience, contact, and construal: How people’s personality traits predict characteristics of their situations in daily life. Journal of Research in Personality, 55, 98–111.

50.

Robinson

M. D.

Ode

Moeller

S. K.

Goetz

P. W.

(2007). Neuroticism and affective priming: Evidence for a neuroticism-linked negative schema. Personality and Individuals Differences, 42, 1221–1231.

51.

Serfass

D. G.

Sherman

R. A.

(2013). Personality and the perceptions of situations from the Thematic Apperception Test. Journal of Research in Personality, 47, 708–718.

52.

Serfass

D. S.

Sherman

R. A.

(2015). Situations in 140 characters: Assessing real-world situations on Twitter. PLoS One, 10, 1–19.

53.

Sherman

R. A.

Nave

C. N.

Funder

D. C.

(2010). Situational similarity and personality predict behavioral consistency. Journal of Personality and Social Psychology, 99, 330–343.

54.

Sherman

R. A.

Nave

C. S.

Funder

D. C.

(2012). Properties of persons and situations related to overall and distinctive personality-behavior congruence. Journal of Research in Personality, 46, 87–101.

55.

Sherman

R. A.

Nave

C. S.

Funder

D. C.

(2013). Situational construal is related to personality and gender. Journal of Research in Personality, 47, 1–14.

56.

Sherman

R. A.

Rauthmann

J. F.

Brown

N. A.

Serfass

D. S.

Jones

A. B.

(2015). The independent effects of personality and situations on real-time expressions of behavior and emotion. Journal of Personality and Social Psychology, 109, 872–888.

57.

van Heck

G. L.

(1984). The construction of a general taxonomy of situations. In Bonarius

Van Heck

G. L.

Smid

(Eds.), Personality psychology in Europe: Theoretical and empirical developments (pp. 149–164). Lisse, the Netherlands: Swets and Zeitlinger.

58.

van Heck

G. L.

(1989). Situation concepts: Definitions and classification. In Hettema

P. J.

(Ed.), Personality and environment: Assessment of human adaptation (pp. 53–69). Oxford, England: Wiley.

59.

Wagerman

S. A.

Funder

D. C.

(2009). Situations. In Corr

P. J.

Matthews

(Eds.), Cambridge handbook of personality psychology (pp. 27–42). Cambridge, MA: Cambridge University Press.

60.

Wells

G. L.

Windschitl

P. D.

(1999). Stimulus sampling and social psychological experimentation. Personality and Social Psychology Bulletin, 25, 1115–1125.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.14 MB