presentation

RESEARCH REPORT

Morning Employees Are Perceived as Better Employees: Employees’ Start Times Influence Supervisor Performance Ratings

Kai Chi Yam, Ryan Fehr, and Christopher M. Barnes University of Washington

In this research, we draw from the stereotyping literature to suggest that supervisor ratings of job performance are affected by employees’ start times—the time of day they first arrive at work. Even when accounting for total work hours, objective job performance, and employees’ self-ratings of conscien- tiousness, we find that a later start time leads supervisors to perceive employees as less conscientious. These perceptions in turn cause supervisors to rate employees as lower performers. In addition, we show that supervisor chronotype acts as a boundary condition of the mediated model. Supervisors who prefer eveningness (i.e., owls) are less likely to hold negative stereotypes of employees with late start times than supervisors who prefer morningness (i.e., larks). Taken together, our results suggest that supervisor ratings of job performance are susceptible to stereotypic beliefs based on employees’ start times.

Keywords: stereotyping, implicit bias, job performance, morningness, chronotype

For most of the 20th century, employees’ work schedules were highly regimented. Office managers and factory workers alike arrived at work at set times determined by their employers, with little say in their schedules and little variability across employees. Today, technological and social forces have provided employees with greater say and more flexibility in when they start and end their work days through programs collectively referred to as flex- ible work practices (FWPs; Kelly & Moen, 2007). FWPs allow employees to meet obligatory duties such as driving their children to school and caring for aging parents, while enabling organiza- tions to attract and retain talent by enhancing employees’ job satisfaction and commitment to the organization.

Meta-analyses suggest that FWPs generally produce desirable individual and organizational outcomes such as increased produc- tivity, higher job satisfaction, and decreased turnover intentions (Baltes, Briggs, Huff, Wright, & Neuman, 1999; Gajendran & Harrison, 2007). However, research suggests that FWPs can also be harmful to employees’ career outcomes (Glass, 2004). For instance, when supervisors believe that employees are only using FWPs to help them manage their personal lives, FWPs have a negative impact on employees’ career success (Leslie, Manchester,

Park, & Mehng, 2012). Although scholars have demonstrated the potential downsides of FWPs, research to date has only begun to consider the precise mechanisms and boundary conditions of these effects. In addition, research has tended to consider FWPs holis- tically, providing limited insight into the specific aspects of FWPs that might be most directly responsible for their negative impact on employees’ success.

In this article, we address these limitations by focusing on employee start times— one of the most frequently discussed and utilized components of FWPs (Galinsky, Bond, & Sakai, 2008). Drawing from the stereotyping literature, we specifically examine the implications of a stereotypic negative perception of people who begin the day’s activities late—referred to here as a morning bias—for employees and their career success. We suggest that supervisors exhibit a pervasive morning bias and stereotype em- ployees with late start times as less conscientious than employees with early start times. These perceptions in turn lead to lower performance ratings for employees with late start times. Further- more, we explore supervisor chronotype as an important boundary condition of the mediated model. Based on decades of research on social identity theory and ingroup favoritism (Tajfel & Turner, 1986), we argue that supervisors with a preference for eveningness exhibit less negative stereotyping of employees with late schedules than supervisors with a preference for morningness, thereby avoid- ing the morning bias when evaluating their employees’ perfor- mance. In other words, we propose a first-stage moderated medi- ation model in which supervisors’ chronotypes act as a boundary condition to affect the link between employees’ start times and their perceived levels of conscientiousness, which in turn affects supervisors’ ratings of employees’ performance (see Figure 1).

By developing and testing a model of supervisors’ automatic inferences about employee job performance, we make several key contributions to the literature. First, we provide insight into the

This article was published Online First June 9, 2014. Kai Chi Yam, Ryan Fehr, and Christopher M. Barnes, Department of

Management and Organization, Michael G. Foster School of Business, University of Washington.

We thank associate editor Deidra Schleicher and participants in the IdeaLab, a student-run research seminar at the Foster School of Business, for constructive comments on earlier drafts of this article.

Correspondence concerning this article should be addressed to Kai Chi Yam, Department of Management and Organization, Michael G. Foster School of Business, University of Washington, Box 353226, Seattle, WA 98195-3226. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Journal of Applied Psychology © 2014 American Psychological Association 2014, Vol. 99, No. 6, 1288 –1299 0021-9010/14/$12.00 http://dx.doi.org/10.1037/a0037109

1288

question of why supervisors might provide differential perfor- mance ratings of employees with equal performance metrics. Sec- ond, we contribute to a growing literature on the role of automatic inferences on employee performance evaluations, demonstrating that these inferences are based on more than gender and racial stereotypes (Jost et al., 2009). Finally, we pinpoint an important dark side to the deployment and utilization of flexible work prac- tices, demonstrating that employees who use FWPs to arrive at work later in the day are likely to receive worse performance evaluations than their equally performing peers.

Stereotyping and Employees’ Start Times

Stereotypes are “beliefs about the characteristics, attributes, and behaviors of members of certain groups” (Hilton & von Hippel, 1996, p. 240). Although stereotypes are not necessarily erroneous, scholars have suggested that a stereotypic belief is often “a fixed impression which conforms very little to the facts it pretends to represent” (Katz & Braly, 1935, p. 181). Within the organizational sciences, scholars have demonstrated that workplace stereotypes are pervasive, particularly with respect to race and gender. Among racial minorities, racial stereotypes negatively influence the like- lihood that they will receive job offers (Bertrand & Mullainathan, 2004) and succeed after a job is obtained (Johnson & Eby, 2011). Among women, gender stereotypes negatively affect hireability ratings of female professors (Madera, Hebl, & Martin, 2009) and hinder women’s decisions to become entrepreneurs (Gupta, Tur- ban, & Bhawe, 2008).

Whereas previous research has primarily focused on racial and gender stereotypes, we suggest that employees’ start times may also be a source of stereotypic beliefs with important work-related consequences. Our contention is that, with all else being equal, a pervasive morning bias causes supervisors to stereotype employ- ees who start work late in the day as less conscientious than employees who start work early in the day. The association be- tween morningness— beginning the day’s activities early in the morning—and goodness is ubiquitous in our society. The old saying “the early bird gets the worm” suggests that people who start their days early in the morning are the most successful. Benjamin Franklin (1855/2014) is famous for his belief that “early to bed, early to rise, makes a man healthy, wealthy and wise,” and Aristotle (1995) mused that “it is well to be up before daybreak, for such habits contribute to health, wealth, and wisdom.” An old Chinese saying similarly advises that “a day’s planning should be done in the morning,” suggesting that the positive perception of morningness transcends cultural boundaries.

Empirical research suggests that the stereotypic association be- tween morningness and goodness can be traced in part to a link between morningness and conscientiousness—a personality trait associated with hard work, self-discipline, and dutifulness (Costa & McCrae, 1992; Kanfer & Heggestad, 1997). Specifically, re- search suggests that people who start the day’s activities late are generally perceived as less conscientious than people who start the day’s activities early (Hepburn, Ortiz, & Locksley, 1984; Mc- Cutcheon, 1998) and that there is a small but positive correlation between self-reports of morningness and conscientiousness (DeY- oung, Hasher, Djikic, Criger, & Peterson, 2007; Hogben, Ellis, Archer, & von Schantz, 2007).

Although employees with preferences for eveningness might be objectively less conscientious than employees with preferences for morningness, we suggest that supervisors’ morning biases equates to an overemphasis on employee start times in consci- entiousness evaluations, leading them to rate employees with late start times as less conscientious than employees with early start times. In various national surveys of FWPs, employees cite family demands as the top reason they utilize FWPs and as the main driver of their start time decisions (Galinsky, Bond, & Hill, 2004; Galinsky et al., 2008). For example, a sales associate who is a morning person may nonetheless choose to start work at 1 p.m. because he needs to attend classes in the morning. Likewise, a parent who is an evening person may nonetheless choose to start work at 7 a.m. because she needs to pick up her child from school in the afternoon. Thus, although employees’ start times are a primary source of information available to supervisors regarding employees’ preferences for morningness versus eveningness, this information is often independent to employees’ actual preferences, and hence can be theorized to influence supervisors’ perceptions of employee conscientiousness above and beyond their actual levels of conscientiousness. We therefore posit the following hypothesis:

Hypothesis 1: Employee start time is negatively associated with supervisor perceptions of employee conscientiousness.

Implications for Supervisor Ratings of Job Performance

We argue that once supervisors stereotype employees as low in conscientiousness based on their start times, these perceptions in turn lead supervisors to rate employees with late start times as lower performers. Lexical studies suggest that conscientious em- ployees are reliable, orderly, and industrious, and have high im- pulse control (Hofstee, de Raad, & Goldberg, 1992; Roberts,

Employee Start Time

Supervisor Perception of

Employee Conscientiousness

Supervisor Chronotype

Supervisor Rating of Employee Job Performance

Figure 1. Theoretical model of the impact of employee start time on supervisor performance ratings.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1289PERCEPTION AND MORNING EMPLOYEES

Bogg, Walton, Chernyshenko, & Stark, 2004), all of which are imperative for high job performance. Supervisors believe that conscientiousness is a precursor to high job performance because conscientious employees display higher levels of work motivation (Colquitt, LePine, & Noe, 2000). In contrast, because supervisors perceive employees who “choose” to start working late as lacking these important personal characteristics, they are rated as lower performers.

Most research to date on the link between conscientiousness and job performance has focused on employees’ self-report conscien- tiousness, rather than supervisor perceptions (Barrick & Mount, 1991; Salgado, 1997; Tett, Jackson, & Rothstein, 1991). However, recent studies also provide evidence consistent with the notion of a positive relationship between supervisors’ perceptions of em- ployee conscientiousness and job performance. In a survey-based study, interviewers were more likely to recommend hiring candi- dates whom they perceived to be high on conscientiousness (van Dam, 2003). In another study, Emerson and Yang (2012) found that perceived levels of conscientiousness in accountants were positively associated with their perceived ability to perform their jobs well. Taken together, these empirical studies suggest a posi- tive association between supervisors’ perceptions of their employ- ees’ conscientiousness and their assessments of these employees’ job performance, leading to the following hypothesis:

Hypothesis 2: The relationship between employee start time and supervisor ratings of job performance is mediated by supervisor perceptions of employee conscientiousness.

The Role of Supervisor Chronotype

Although the morning bias is pervasive in our society, we suggest that the stereotypic belief that start times influence employee-perceived conscientiousness depends in part on super- visors’ own chronotypes. Chronotypes are “patterns of sleep/wake activity and energy levels that are tied to time of day and governed by internal circadian and sleep drives” (Biss & Hasher, 2012, p. 437). People can be classified as either “larks,” who wake up and go to bed early, or “owls,” who wake up and go to bed late (Horne & Ostberg, 1976). Drawing from social identity theory (SIT; Taifel & Turner, 1986), we suggest that supervisors who are themselves evening people are less susceptible to the stereotypic belief that employees who work late are less conscientious and worse per- formers than supervisors who are morning people.

According to SIT, people tend to spontaneously self-categorize themselves into various social groups. Although such categoriza- tions help people make sense of their social environments by providing prototypes for different group memberships and allow- ing people to act in accordance with those norms, they can also lead to ingroup favoritism (Terry & Hogg, 1996). SIT posits that people are motivated to evaluate their own social groups more positively in order to maintain a positive social identity and self- esteem (Tajfel, 1978). Social identities often develop as a function of highly visible group memberships (e.g., job function; Ashforth & Mael, 1989) but can also develop as a function of differences as subtle as musical taste (Lonsdale & North, 2009). Consistent with this literature, we argue that supervisors who are owls are likely to self-categorize themselves with others whom they perceive to prefer eveningness in the workplace. Once this social identity is

formed, supervisors who prefer eveningness are less likely than their peers to exhibit a negative morning bias because they per- ceive themselves as belonging to the “owls” group. These argu- ments suggest the following hypothesis:

Hypothesis 3: The indirect effect of employee start time on supervisor ratings of performance via perceived conscien- tiousness is moderated by supervisors’ chronotypes, such that the indirect effect will be strong when supervisors are “larks,” but dissipate when supervisors are “owls.”

Research Overview

We conducted three studies to test our theoretical model. In Study 1, we used an experimental method to examine the impact of morningness cues on the accessibility of words related to conscientiousness. In Study 2, we then tested our full theoretical model by surveying supervisor– employee dyads and examined the effect of employee start time on supervisor perceptions of consci- entiousness and job performance, controlling for total work hours, supervisor– employee relationship quality, and self-rated consci- entiousness. Finally, in Study 3, we constructively replicated our findings with a randomized performance rating exercise to estab- lish causal inferences for our theoretical model.

Study 1

Method

We recruited 120 working adults (Mage � 29.8 years, 76.1% European American; 58.7% female) from Amazon’s Mechanical Turk. Following from the conceptualization of stereotypes as im- plicit and automatic biases (Blair, 2002), we began our empirical examination of the hypothesized model with a lexical decision task (LDT) to assess the automatic association between morningness and conscientiousness. The LDT is an association-based implicit measure that examines the accessibility of a concept in a person’s mind by examining how quickly he or she can distinguish words associated with a focal concept from nonwords as a function of a pretask prime (Uhlmann et al., 2012). For example, in past re- search, the LDT has been used to establish a link between a leader’s gender and perceptions of his or her communal leadership traits by priming participants to think of either a female or a male leader and then assessing how quickly they are able to recognize words indicative of communal leadership traits (Scott & Brown, 2006). In these studies, a stronger association between two con- cepts (e.g., a female leader and communal leadership traits) equates to faster reaction times because exposure to the first concept during the priming phase of the experiment enhances the mental accessibility of the second concept during the decision- making phase. In contrast, an inverse relationship between two concepts should produce slower reaction times (Bassili, 2003).

Participants were randomly assigned to one of two priming conditions. In both conditions, we presented participants with 10 word fragments and asked them to complete the fragments to form valid English words. In the morningness [eveningness] condition, five fragments were related to morningness (e.g., sunrise) [eve- ningness (e.g., sunset)]. We also embedded five neutral fragments in both conditions (e.g., elephant) to conceal the purpose of the

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1290 YAM, FEHR, AND BARNES

study. On average, participants successfully completed 4.20 out of the five morningness or eveningness fragments, indicating that participants were attending to the stimuli and correctly following the instructions.

After the prime, we directed participants to complete the LDT. Participants were randomly presented with strings of letters and asked to quickly identify, using keystrokes of E or I, whether the strings constituted valid English words or not. Participants were presented with 10 nonsense strings (pronounceable nonwords), 10 neutral words (e.g., notebook), and 10 conscientiousness-related words in random order. The 10 conscientiousness-related words were chosen from extant lexical investigations of conscientiousness, which specify five unique facets of conscientiousness: orderliness, reliability, industriousness, impulse control, and decisiveness (Roberts et al., 2004). Specifically, we chose the two positive- valence words with the highest factor loadings from each facet (i.e., organized, neat, reliable, dependable, industrious, tenacious, careful, cautious, decisive, and deliberate; Roberts et al., 2004). Participants were instructed to sort real words from nonwords as quickly as possible. Consistent with prior research, we excluded incorrect responses and reaction times that exceeded 3 s, as these responses may be a result of participants’ temporary lack of attention to the task (Greenwald, Nosek, & Banaji, 2003). A total of 11 participants were omitted due to either a high rate of incorrect responses (indicating randomly clicking) or significant incomplete data.

Results and Discussion

To test Hypothesis 1, we examined the impact of exposure to morningness versus eveningness cues on participants’ averaged reaction times (in milliseconds) to conscientiousness-related words. According to Hypothesis 1, participants in the morningness condition should respond to conscientiousness-related words more quickly than participants in the eveningness condition. As ex- pected, participants in the morningness condition (M � 394.38, SD � 152.12) reacted to the conscientiousness words significantly faster than participants in the eveningness condition (M � 465.83, SD � 180.39), t(107) � �2.24, p � .05. Paired-samples t tests further showed that participants in the morningness condition reacted to the conscientiousness words significantly faster than the neutral words (M � 433.65, SD � 159.35), t(54) � �2.04, p � .05, and that participants in the eveningness condition reacted to the conscientiousness words significantly slower than the neutral words (M � 427.09, SD � 172.44), t(53) � 3.14, p � .01. Reaction times for the neutral words did not differ by condition, t(107) � �0.16, p � .88. These results provide initial support for Hypothesis 1, suggesting that participants display a morning bias and automatically associate morningness with conscientiousness. On the basis of this initial supporting evidence of a morningness bias, we conducted Study 2 to test our full theoretical model in an organizational setting (i.e., first-stage moderated mediation).

Study 2

Method

Participants and procedures. We recruited 229 supervisor– employee dyads through the Study Response Project (for a detailed

discussion of Study Response, see www.studyresponse.net; for recent examples of studies utilizing the Study Response data collection method for dyadic data, see Barnes, Schaubroeck, Huth, & Ghumman, 2011; Piccolo & Colquitt, 2006). With the assistance of the Study Response administrators, we first prescreened poten- tial participants to select those who were working full-time and willing to invite their supervisors to participate in a study on workplace attitudes. Administrators from the Study Response Proj- ect then validated all leaders’ e-mail addresses. A total of 149 dyads successfully completed the surveys (65.1% response rate). Employees were an average of 39.23 years old, 62.4% male, and 79.6% European American. Supervisors were an average of 41.98 years old, 70.7% male, and 85.3% European American. Employees completed a measure of the time that they arrive at work. Super- visors completed a measure of their own chronotypes and rated their employees’ conscientiousness and job performance. Employ- ees and supervisors also completed a set of measures for control variables.

Measures. Employee start time. We measured employees’ start times by

asking employees, “On average, when do you come to work?” Participants were asked to provide a specific time. Employees’ start times ranged from 5 a.m. to 9:45 a.m. with a mean start time of 8:42 a.m. We standardized this variable to ease interpretation of the data.

Supervisor perceptions of employee conscientiousness. We asked supervisors to rate their employees’ conscientiousness with a 10-item Conscientiousness Scale (e.g., “This employee is always prepared”; Goldberg et al., 2006; 1 � strongly disagree; 5 � strongly agree; � � .88).

Supervisor ratings of employee job performance. We asked supervisors to rate their employees’ job performance with a five- item Job Performance Scale (e.g., “This employee always com- pletes the duties specified in his or her job description”; Podsakoff & MacKenize, 1989; 1 � strongly disagree; 7 � strongly agree; � � .81).

Supervisor chronotype. Supervisors’ chronotypes were mea- sured via the 12-item Morning–Evening Orientation Scale (Smith et al., 2002). Supervisors were asked to respond to the questions according to their own preferences and not what they may be forced to do by their work schedules (e.g., “When would you prefer to get up if you had a day off and nothing to do?”; 1 � much later than most people; 5 � much earlier than most people; � � .90). A higher score indicates a greater preference for morning- ness.

Control variables. In addition to demographic information, we controlled for employees’ total work hours because perfor- mance evaluations are likely affected by total hours worked. Em- ployees were asked to indicate their total work hours at home and at the office in a typical week. Supervisors were asked to estimate employees’ total work hours with the same items. We added these items together to form two total work hours variables (employee- reported and supervisor’s perceptions). Employees self-reported their conscientiousness with an identical 10-item scale. We con- trolled for self-report conscientiousness so that our results are not confounded by actual differences in conscientiousness across em- ployees. We also controlled for whether flexibility in start times is a formal arrangement between the dyad with one item (“My manager gives me flexibility over my work schedule”; 1 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1291PERCEPTION AND MORNING EMPLOYEES

strongly disagree; 5 � strongly agree). For supervisors, we con- trolled for relative start time by asking supervisors to indicate whether the focal employees come to work earlier or later than other people with similar jobs (1 � much earlier; 5 � much later) to control for social comparison effects. Finally, we controlled for the length and quality of interaction between the employee and the supervisor because ratings of others’ personalities are affected by familiarity with the targets (Connelly & Ones, 2010). Supervisors were asked how long they have worked with their employees (in years), and how well they know their employees (1 � not at all, 5 � very well).1

Results and Discussion

The means, standard deviations, and correlations of the focal variables are presented in Table 1. Prior to hypothesis testing, we first conducted a confirmatory factor analysis (CFA) to demon- strate discriminant validity among our three latent constructs: supervisors’ ratings of employees’ job performance, supervisors’ perceptions of employees’ conscientiousness, and supervisors’ chronotypes. The hypothesized three-factor model demonstrated good fit to the data, �2(323) � 656.60, p � .01, comparative fit index (CFI) � .91, root-mean-square error of approximation (RMSEA) � .08, and was also superior to a more parsimonious model where the conscientiousness and job performance items were set to load on a single factor, ��2(1) � 7.41, p � .01.

To test Hypothesis 1, we conducted an ordinary least squares (OLS) regression. All control variables were entered in Model 1 to predict supervisor perceptions of employees’ conscientiousness. In Model 2, we entered employees’ standardized start times as a predictor and found that employees who come to work later were perceived as less conscientious by their supervisors (� � �.18, p � .05; see Table 2). We then conducted a bias-corrected boot- strapping procedure (1,000 resamples) to test for mediation (Hy- pothesis 2) because traditional methods (e.g., Baron & Kenny, 1986) in testing mediation are generally low in power (Fritz & MacKinnon, 2007). The coefficient for the indirect effect of em- ployees’ start times on supervisor ratings of job performance via

supervisor perceptions of conscientiousness was significant (indi- rect effect � �.09, SE � .04, 95% confidence interval [CI] [�.23, �.01]). Collectively, these results provide support for Hy- pothesis 1 and Hypothesis 2, suggesting that, all else being equal, employees with late start times are rated as lower performers by their supervisors than employees with early start times, mediated by reduced perceptions of conscientiousness.

To test our hypothesized first-stage moderated mediation model, we began by examining the interactive effect of supervisor chro- notype and employee start time on supervisor perceptions of conscientiousness. We entered supervisor chronotype and em- ployee start time in Model 2 and the interaction term in Model 3. Results suggested that the model explained significantly more variance (�R2 � .02, p � .05) and that the interaction term was significant (� � �.38, p � .01; see Table 2). Simple slope analysis confirmed that results were in the expected direction. When su- pervisors were “larks,” employee start time significantly predicted lower levels of perceived conscientiousness (� � �.44, p � .01). When supervisors were “owls,” however, employee start time did not influence perceptions of conscientiousness as the slope did not differ significantly from zero (� � �.18, p � .14). To aid interpretation, we have plotted the interaction effect in Figure 2.

Finally, we utilized the methods of Hayes (2013) to test for conditional indirect effects at 1 standard deviation above and below the mean of the moderator (i.e., supervisor chronotype). When supervisors were larks, the mediated model was significant (indirect effect � �.14, SE � .07, 95%, CI [�.30, �.01]. When

1 Because we did not collect data on employees’ chronotypes, we conducted a validation study to provide support for the argument that employees’ start times and their chronotypes are uncorrelated constructs. We recruited 135 working adults (six were dropped due to significant missing data) from Amazon’s Mechanical Turk (Mage � 32.16 years, 79.1% European American; 50.4% female). On average, these participants reported 12.43 (SD � 8.31) years of working experience. We used the same measures to assess participants’ start times and chronotypes (higher � “larks”) as in Study 2. As expected, start time and chronotype were not significantly correlated (r � �.08, p � .41).

Table 1 Means, Standard Deviations, and Correlations of the Focal Variables (Study 2)

Variable M (SD) 1 2 3 4 5 6 7 8 9 10 11

Employee-report variables 1. Employee start timea,d 0.00 (1.00) (—) 2. Total work hoursd 0.00 (1.00) .06 (—) 3. Formal arrangement 4.94 (1.64) .17� .18� (—) 4. Self-report conscientiousness 3.91 (0.65) �.12 .03 .02 (.84)

Supervisor-report variables 5. Perceptions of conscientiousness 3.83 (0.78) �.26�� .03 .05 .23�� (.88) 6. Job performance 5.68 (0.93) �.10 �.04 .12 .24�� .55�� (.81) 7. Supervisor chronotypeb 3.26 (.60) �.12 .10 .12 �.10 �.14† �.15† (.90) 8. Total work hoursd 0.00 (1.00) �.23�� .49�� .11 .09 .03 .04 .07 (—) 9. Dyad interaction (years) 4.30 (1.05) �.02 .03 .03 .06 .11 .07 .09 .18� (—) 10. Dyad relationship quality 4.08 (0.78) .02 �.07 .13† �.02 .32�� .35�� .02 .02 .10 (—) 11. Relative start timec 2.40 (0.74) .25�� �.04 .18� �.00 �.29�� �.11 .04 �.23�� �.07 .03 (—)

Note. N � 149. Alpha coefficients are presented on the diagonal. a Higher scores indicated later start hours. b Higher scores indicated greater preferences for morningness. c Higher scores indicated later relative start time. d Standardized scores. † p � .10. � p � .05. �� p � .01.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1292 YAM, FEHR, AND BARNES

supervisors were owls, however, the mediated model was not significant (indirect effect � �.02, SE � 0.10, 95% CI[�.22, .16]). These results hold even when we account for all the control variables. Together, the results suggest that supervisors who prefer eveningness are less likely to be affected by the morning bias and less likely to rate employees with late start times as less consci- entious or lower performers.

Study 2 provides support for Hypotheses 1, 2, and 3, demon- strating that employee start times are meaningfully related to supervisors’ performance ratings. Despite its strengths, there are two key limitations to Study 2. First, although high in external validity, the cross-sectional design of Study 2 precluded us from establishing causal inferences. Second, although Study 2 con- trolled for a wide array of factors, we were unable to collect comparable data on objective performance and to rule out this potential confound. In Study 3, we addressed these concerns with

an experimental design, allowing us to establish causality and hold objective performance constant through the use of a simulated performance rating task.

Study 3

Method

Participants and procedure. We recruited 150 undergraduate students (Mage � 21.11 years; 57.4% female; 55.3% European Amer- ican) from a large public university to complete a performance rating task. A total of 85% reported at least 1 year of work experience. Nine participants were dropped due to significant missing data, leaving a final sample of 141 students. We adapted a previously validated performance rating task in which participants were asked to assume the role of a manager at a fictitious company and to rate one of their employees (Reb & Cropanzano, 2007). Participants were randomly assigned to one of two conditions. In the early [late] start time condition, participants read about the performance of an employee who comes to work early [late] in the day. After reading the employee description, participants were asked to rate the employee in terms of his or her conscientiousness and job performance. Participants then completed a measure of chronotype and provided demographic infor- mation.

Manipulation and measures. Employee start time manipulation. We randomly presented

participants with one of the two profiles. In the early [late] start time condition, participants were told that, in general, the em- ployee comes to work at 7 a.m. [11 a.m.] and leaves at 3 p.m. [7 p.m.] (see Appendix). The employee’s schedule was approved by the company, and in both conditions, the employee worked for 8 hr per day. In addition, we presented participants with an identical objective performance chart of the employee over the last 26 weeks to hold objective performance constant.

Table 2 Summary of Regression Results (Study 2)

Variable

Dependent variable � Supervisor perceptions of conscientiousness

Model 1 Model 2 Model 3

B SE B � B SE B � B SE B �

Age .03 .01 .34�� .03 .01 .34�� .03 .01 .32��

Racea .09 .13 .05 .08 .13 .05 .11 .13 .06 Genderb .12 .06 .16� .13 .06 .17� .13 .06 .17�

Employee total work hours (supervisor report) �.01 .07 �.01 �.02 .08 �.03 �.04 .09 �.04 Self-reported conscientiousness .20 .09 .17� .19 .09 .16� .19 .10 .16�

Dyad relationship quality .37 .08 .38�� .37 .08 .38�� .35 .08 .36��

Dyad interaction (years) .11 .06 .16† .12 .06 .17� .12 .06 .18�

Employee total work hours (employee report) .06 .07 .08 .07 .07 .09 .08 .08 .11 Formal arrangement .03 .04 .02 .03 .04 .02 .03 .04 .03 Relative start time �.06 .06 �.08 �.04 .09 �.04 �.06 .09 �.06 Employee start time �.16 .06 �.18� �.18 .04 �.25��

Supervisor chronotype �.08 .10 �.06 �.16 .10 �.15†

Employee Start Time Supervisor Chronotype �.26 .10 �.38��

Adjusted R2 .30 .32 .34 �R2 .02� .02�

Note. N � 149. a Dummy variable (1 � White, 0 � others). b Dummy variable (1 � female, 0 � male). † p � .10. � p � .05. �� p � .01.

1

2

3

4

5

Early Start Time Late Start Time

Su pe

rv is

or P

er ce

pt io

ns o

f C

on sc

ie nt

io su

ne ss

Manager = Owl

Manager = Lark

Figure 2. The interactive effect of employees’ start times and supervi- sors’ chronotypes on supervisors’ perceptions of employees’ conscien- tiousness (Study 2).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1293PERCEPTION AND MORNING EMPLOYEES

Supervisor perceptions of employee conscientiousness. We asked participants to rate the employee’s conscientiousness with the same 10-item Conscientiousness Scale as in Study 2 (Goldberg et al., 2006; 1 � strongly disagree; 5 � strongly agree; � � .90; M � 2.94; SD � 0.70).

Supervisor ratings of job performance. We asked participants to rate the employee’s job performance with the same five-item Job Performance Scale as in Study 2 (Podsakoff & MacKenize, 1989; 1 � strongly disagree; 7 � strongly agree; � � .74; M � 4.74; SD � 0.73).

Supervisor chronotype. We asked participants to rate their own chronotypes with the same 12-item morning– evening orien- tation scale as in Study 2 (Smith et al., 2002; 1 � much later than most people; 5 � much earlier than most people; � � .82; M � 2.77; SD � 0.60).

Results and Discussion

The means, standard deviations, and correlations of the focal variables are presented in Table 3. Prior to hypothesis testing, we first conducted a CFA. The hypothesized three-factor model dem- onstrated good fit to the data, �2(323) � 401.16, p � .01, CFI � .92, RMSEA � .06, and was also superior to a more parsimonious model where the conscientiousness and job performance items were set to load on a single factor (��[1

2] � 90.34, p � .01). To test Hypothesis 1, we conducted an independent samples t

test. Participants in the late start time condition rated the employee as significantly less conscientious (M � 2.83, SD � 0.65) than participants in the early start time condition (M � 3.06, SD � 0.73), t(139) � �2.01, p � .05. We then conducted a bias- corrected bootstrapping procedure (1,000 resamples) to test for mediation (Hypothesis 2) as in Study 2. The coefficient for the indirect effect of employee start time on participants’ ratings of job performance via perceptions of conscientiousness was significant (indirect effect � �.04, SE � .03, 95% CI [�.13, �.003]). These results support Hypotheses 1 and 2.

As in Study 2, we examined Hypothesis 3 by first assessing the interactive effect of participants’ chronotypes and employee start time on participants’ perceptions of the employee’s conscientious- ness. Chronotype and the experimental condition were entered in Model 1 of the OLS regression, and the interaction term was entered in Model 2. Results suggested that the model explained significantly more variance (�R2 � .03, p � .05) and that the interaction effect was significant (� � �.17, p � .05; see Table 4). Simple slope analysis confirmed that when participants preferred morningness, employee start time significantly predicted lower

levels of perceived conscientiousness (� � �.21, p � .05). When participants preferred eveningness, however, employee start time did not influence perceptions of conscientiousness as the slope did not differ significantly from zero (� � �.07, p � .49). To aid interpretation, we have plotted the interaction effect in Figure 3. We then utilized the methods of Hayes (2013) to test for condi- tional indirect effects. When participants were larks, the mediated model was significant (indirect effect � �.06, SE � 0.05, 95% CI [�.20, .004]). When participants were owls, however, the medi- ated model was not significant (indirect effect � �.03, SE � 0.04, 95% CI [�.16, .03]). Together, the results provide support for our full theoretical model.

General Discussion

Contemporary employees are increasingly concerned about work–life balance. To this end, many organizations have intro- duced flexible work practices that allow employees to decide when to start the work day. Drawing from the stereotyping literature, we conducted three studies aimed at providing insight into the pro- cesses and boundary conditions that underlie the negative effects of FWPs on how employees are perceived. Specifically, we dem- onstrated that employees who choose late start times subject them- selves to the negative stereotype of being less conscientious. This stereotype in turn leads supervisors to perceive late-starting em- ployees as lower performers, but only among supervisors who prefer morningness themselves.

Theoretical Implications

First and foremost, we contribute to the FWPs literature by introducing stereotyping as an important lens through which the negative effects of FWPs can be examined. Popular beliefs and empirical research suggest that FWPs are beneficial for employees (e.g., Baltes et al., 1999).

Nonetheless, recent studies have begun to challenge this view, suggesting a potential dark side to FWPs (Glass, 2004; Leslie et al., 2012). Through the lens of the stereotyping literature, we provide insight into the question of why, precisely, FWPs might produce adverse effects. Our results suggest that employees should be particularly mindful of their decisions to utilize FWPs by arriving at work later in the day, as this decision might equate to career penalties in the form of lower supervisor performance ratings.

More broadly, our research contributes to the literature by examining the effects of automatic inferences and social percep-

Table 3 Means, Standard Deviations, and Correlations of the Focal Variables (Study 3)

Variable Means (SD) 1 2 3 4

1. Employee start timea 0.50 (0.50) (—) 2. Perceptions of conscientiousness 2.94 (0.70) �.17� (.90) 3. Job performance 4.74 (0.73) �.29�� .22�� (.74) 4. Supervisor chronotypeb 2.77 (0.60) .01 �.03 �.10 (.82)

Note. N � 141. Alpha coefficients are presented on the diagonal. a Dummy variable (0 � early start time; 1 � late start time). b Higher scores indicated greater preferences for morningness. � p � .05. �� p � .01.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1294 YAM, FEHR, AND BARNES

tions in the workplace beyond gender and racial stereotypes. Although decades of research have demonstrated the consequences of gender and racial stereotypes in the workplace, more recent research suggests that even extraneous attributes such as height (Judge & Cable, 2004) and weight (Judge & Cable, 2011) can affect employees’ career success. In addition to these extant work- place stereotypes, we introduce employee start time as another important source of stereotyping that has significant work-related consequences. In addition to introducing a new source of work- place stereotypes, we provide a deeper understanding of when and why this stereotype is most likely to occur by demonstrating the underlying psychological mechanism as well as the boundary conditions of this morning bias. Furthermore, our findings support the contention that stereotyping is a malleable process by revealing that supervisors who are “owls” are least affected by this morning bias (Blair, 2002).

Finally, we contribute to the literature by introducing the con- struct of chronotype in the prediction of supervisor ratings of job performance. Chronotype is a biologically based individual differ- ence that has received consistent empirical attention in the biolog- ical sciences (e.g., Roenneberg et al., 2004) and personality psy- chology (e.g., Tsaousis, 2010) but comparatively little attention in the organizational sciences. In this research, we demonstrate the benefits of applying this construct toward a better understanding of

interpersonal interactions at work. Beyond research on employee start times, we suggest that chronotype can be used to augment research on the role of other individual traits (e.g., Big Five) in determining how people behave at work. For example, a mismatch of chronotype and work hours may lead to depletion and nega- tively affect a range of important work outcomes such as ethical behavior (Barnes et al., 2011) and work effort (Hagger, Wood, Stiff, & Chatzisarantis, 2010).

Practical Implications

Given recent trends indicating a rapid increase in the utilization of FWPs, our research holds several important practical implica- tions for employees, supervisors, and organizations. First, by dem- onstrating the existence of a morning bias and its work-related consequences, we hope to aid employees in deciding whether to utilize FWPs. We are by no means discouraging employees from utilizing FWPs but merely wish to provide them with an under- standing of the full range of potential consequences of using FWPs. Unlike gender or racial stereotypes, there are no legal regulations concerning the morning bias in the workplace. Never- theless, we hope that our results can help supervisors better un- derstand how they rate their employees’ performance. Should supervisors consistently endorse the morning bias and make unfair reward decisions, their employees are likely to suffer. We recom- mend that supervisors remain mindful of the morning bias and ensure that their performance ratings are based on more reliable performance metrics. When organizations implement FWPs, we specifically recommend that they structure their performance ap- praisal systems in ways that might allow them to remain objective in their assessments of employee performance. For example, top management at Best Buy attempts to set a cultural norm for FWPs by explicitly requesting supervisors to let go of their established expectations of traditional work time (Moen, Kelly, & Hill, 2011). By directly addressing these biases upfront, organizations can ensure a more productive and effective workforce, with perfor- mance appraisal systems that accurately reflect the quality of their employees’ work.

Limitations and Future Research

Looking at the limitations of the current research, we suggest a few fruitful avenues for future research. First, our research is

Table 4 Summary of Regression Results (Study 3)

Variable

Dependent variable � Supervisor perceptions of conscientiousness

Model 1 Model 2

B SE B � B SE B �

Experimental conditiona �0.25 0.11 �.18� �0.09 0.05 �.07 Supervisor chronotype �0.06 0.10 �.03 �0.22 0.13 �.20†

Experimental Condition Supervisor Chronotype �0.14 0.05 �.17�

Adjusted R2 0.09 0.12 �R2 0.03�

a Dummy variable (0 � early start time; 1 � late start time). † p � .10. � p � .05.

1

2

3

4

5

Early Start Time Late Start Time

Su pe

rv is

or P

er ce

pt io

ns

of C

on sc

ie nt

io us

ne ss

Manager = Owl

Manager = Lark

Figure 3. The interactive effect of employees’ start times and supervi- sors’ chronotypes on supervisors’ perceptions of employees’ conscien- tiousness (Study 3).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1295PERCEPTION AND MORNING EMPLOYEES

limited in its focus on global job performance, which is only one of many potential outcomes relevant to the morning bias. Job performance has been conceptualized as a multidimensional con- struct encompassing task, citizenship, and counterproductive per- formance (Rotundo & Sackett, 2002). Conscientiousness is gen- erally found to be a stronger predictor of citizenship behavior than task-related performance (for a review, see Borman, Penner, Allen, & Motowidlo, 2001), suggesting that negative effects of the morn- ing bias may be even more salient if we examine other facets of job performance. Conversely, the morning bias may in fact positively affect ratings of creative performance, as research has generally revealed a negative correlation between conscientiousness and creativity (Feist, 1998). We therefore encourage future research to examine the full range of consequences of the morning bias in the workplace.

Second, in Study 3, we asked participants to evaluate employ- ees’ job performance with limited information. Past research sug- gests that people are more likely to rely on stereotypes when making judgments concerning unfamiliar individuals (Fiske, 1998). Moreover, although participants were provided with the same objective performance data, it is likely that participants would form idiosyncratic interpretation of the performance charts. Thus, it is not clear if the findings from Study 3 would fully extend to a field setting. In future research, investigators should seek to address this issue by controlling for objective performance in field settings.

Third, we have only examined supervisor chronotype as a boundary condition of the theoretical model. Future research should examine the moderating effects of contextual factors such as organizational climate and industry type on the link between employee start times and supervisor performance ratings. Ideally, future research can explore these avenues to identify effective structural interventions that can simultaneously leverage the pos- itive effects of FWPs and minimize their negative effects.

References

Aristotle. (1995). Complete works of Aristotle (J. Barnes, Trans.). Prince- ton, NJ: Princeton University Press.

Ashforth, B., & Mael, F. (1989). Social identity theory and the organiza- tion. Academy of Management Review, 14, 20 –39. doi:10.5465/AMR .1989.4278

Baltes, B., Briggs, T., Huff, J., Wright, J., & Neuman, G. (1999). Flexible and compressed workweek schedules: A meta-analysis of their effects on work-related criteria. Journal of Applied Psychology, 84, 496 –513. doi:10.1037/0021-9010.84.4.496

Barnes, C., Schaubroeck, J., Huth, M., & Ghumman, S. (2011). Lack of sleep and unethical conduct. Organizational Behavior and Human De- cision Processes, 115, 169 –180. doi:10.1016/j.obhdp.2011.01.009

Baron, R. M., & Kenny, D. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi:10.1037/0022-3514.51.6.1173

Barrick, M., & Mount, M. (1991). The Big Five personality dimensions and job performance: A meta-analysis. Personnel Psychology, 44, 1–26. doi:10.1111/j.1744-6570.1991.tb00688.x6570.1991.tb00688.x

Bassili, J. (2003). Cognitive indices of social information processing. In A. Tesser & N. Schwartz (Eds.), The Blackwell handbook of social psy- chology: Intraindividual processes (pp. 68 – 88). Oxford, England: Blackwell.

Bertrand, M., & Mullainathan, S. (2004). Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review, 94, 991–1013. doi:10.1257/ 0002828042002561

Biss, R. K., & Hasher, L. (2012). Happy as a lark: Morning-type younger and older adults are higher in positive affect. Emotion, 12, 437– 441. doi:10.1037/a0027071

Blair, I. (2002). The malleability of automatic stereotypes and prejudice. Personality and Social Psychology Review, 6, 242–261. doi:10.1207/ S15327957PSPR0603_8

Borman, W., Penner, L., Allen, T., & Motowidlo, S. (2001). Personality predictors of citizenship performance. International Journal of Selection and Assessment, 9, 52– 69. doi:10.1111/1468-2389.00163

Connelly, B., & Ones, D. (2010). An other perspective on personality: Meta-analytic integration of observers’ accuracy and predictive validity. Psychological Bulletin, 136, 1092–1122. doi:10.1037/a0021212

Colquitt, J. A., LePine, J., & Noe, R. (2000). Toward an integrative theory of training motivation: A meta-analytic path analysis of 20 years of research. Journal of Applied Psychology, 85, 678 –707. doi:10.1037/ 0021-9010.85.5.678

Costa, P., & McCrae, R. (1992). Four ways five factors are basic. Person- ality and Individual Differences, 13, 653– 665. doi:10.1016/0191- 8869(92)90236-I

DeYoung, C., Hasher, L., Djikic, M., Criger, B., & Peterson, J. (2007). Morning people are stable people: Circadian rhythm and the higher- order factors of the Big Five. Personality and Individual Differences, 43, 267–276. doi:10.1016/j.paid.2006.11.030

Emerson, D., & Yang, L. (2012). Perceptions of auditor conscientiousness and fraud detection. Journal of Forensic & Investigative Accounting, 4, 110 –141.

Feist, G. (1998). A meta-analysis of personality in scientific and artistic creativity. Personality and Social Psychology Review, 2, 290 –309. doi: 10.1207/s15327957pspr0204_5

Fiske, S. (1998). Stereotyping, prejudice, and discrimination. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 357– 411). New York, NY: McGraw-Hill.

Franklin, B. (2014). Early to bed, and early to rise, makes a man healthy, wealthy, and wise, or, early rising, a natural, social, and religious duty. Charleston, SC: Nabu Press. (Original work published 1855)

Fritz, M., & MacKinnon, D. (2007). Required sample size to detect the mediated effect. Psychological Science, 18, 233–239. doi:10.1111/j .1467-9280.2007.01882.x

Gajendran, R., & Harrison, D. (2007). The good, the bad, and the unknown about telecommuting: Meta-analysis of psychological mediators and individual consequences. Journal of Applied Psychology, 92, 1524 – 1541. doi:10.1037/0021-9010.92.6.1524

Galinsky, E., Bond, J., & Hill, E. (2004). When work works: A status report on workplace flexibility. New York, NY: Families and Work Institute.

Galinsky, E., Bond, J., & Sakai, K. (2008). 2008 national study of employ- ers. Retrieved from http://familiesandwork.org/site/research/reports/ 2008nse.pdf

Glass, J. (2004). Blessing or curse? Work–family policies and mother’s wage growth over time. Work and Occupations, 31, 367–394. doi: 10.1177/0730888404266364

Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & Gough, H. C. (2006). The International Personality Item Pool and the future of public-domain personality measures. Journal of Research in Personality, 40, 84 –96. doi:10.1016/j.jrp.2005.08.007

Greenwald, A., Nosek, B., & Banaji, M. (2003). Understanding and using the implicit association test: I. An improved scoring algorithm. Journal of Personality and Social Psychology, 85, 197–216. doi:10.1037/0022- 3514.85.2.197

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1296 YAM, FEHR, AND BARNES

Gupta, V., Turban, D., & Bhawe, N. (2008). The effect of gender stereo- type activation on entrepreneurial intentions. Journal of Applied Psy- chology, 93, 1053–1061. doi:10.1037/0021-9010.93.5.1053

Hagger, M., Wood, C., Stiff, C., & Chatzisarantis, N. (2010). Ego depletion and the strength model of self-control: A meta-analysis. Psychological Bulletin, 136, 495–525. doi:10.1037/a0019486

Hayes, A. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guil- ford Press.

Hepburn, C., Ortiz, V., & Locksley, A. (1984). Morning and evening people: Examination of an identifying scale and of social and self- perceptions of personality differences. Journal of Research in Person- ality, 18, 99 –109. doi:10.1016/0092-6566(84)90041-2

Hilton, J., & von Hippel, W. (1996). Stereotypes. Annual Review of Psychology, 47, 237–271. doi:10.1146/annurev.psych.47.1.237

Hofstee, W., de Raad, B., & Goldberg, L. (1992). Integration of the Big Five and circumplex approaches to trait structure. Journal of Personality and Social Psychology, 63, 146 –163. doi:10.1037/0022-3514.63.1.146

Hogben, A., Ellis, J., Archer, S., & von Schantz, M. (2007). Conscien- tiousness is a predictor of diurnal preference. Chronobiology Interna- tional, 24, 1249 –1254. doi:10.1080/07420520701791596

Horne, J., & Ostberg, O. (1976). A self-assessment questionnaire to deter- mine morningness– eveningness in human circadian rhythms. Interna- tional Journal of Chronobiology, 4, 97–110.

Johnson, C., & Eby, L. (2011). Evaluating career success of African American males: It’s what you know and who you are that matters. Journal of Vocational Behavior, 79, 699 –709. doi:10.1016/j.jvb.2011.03.021

Jost, J., Rudman, L., Blair, I., Carney, D., Dasgupta, N., Glaser, J., & Hardin, C. (2009). The existence of implicit bias is beyond reasonable doubt: A refutation of ideological and methodological objections and executive summary of ten studies that no manager should ignore. Re- search in Organizational Behavior, 29, 39 – 69. doi:10.1016/j.riob.2009 .10.001

Judge, T., & Cable, D. (2004). The effect of physical height on workplace success and income: Preliminary test of a theoretical model. Journal of Applied Psychology, 89, 428 – 441. doi:10.1037/0021-9010.89.3.428

Judge, T., & Cable, D. (2011). When it comes to pay, do the thin win? The effect of weight on pay for men and women. Journal of Applied Psychology, 96, 95–112. doi:10.1037/a0020860

Kanfer, R., & Heggestad, E. (1997). Motivational traits and skills: A person-centered approach to work motivation. Research in Organiza- tional Behavior, 19, 1–57.

Katz, D., & Braly, K. (1935). Racial prejudice and racial stereotypes. Journal of Abnormal and Social Psychology, 30, 175–193. doi:10.1037/ h0059800

Kelly, E., & Moen, P. (2007). Rethinking the clockwork of work: Why schedule control may pay off at work and at home. Advances in Devel- oping Human Resources, 9, 487–506. doi:10.1177/1523422307305489

Leslie, L., Manchester, C., Park, T., & Mehng, S. (2012). Flexible work practices: A source of career premiums or penalties? Academy of Man- agement Journal, 55, 1407–1428. doi:10.5465/amj.2010.0651

Lonsdale, A., & North, A. (2009). Musical taste and ingroup favoritism. Group Processes & Intergroup Relations, 12, 319 –327. doi:10.1177/ 1368430209102842

Madera, J., Hebl, M., & Martin, R. (2009). Gender and letters of recom- mendation for academia: Agentic and communal differences. Journal of Applied Psychology, 94, 1591–1599. doi:10.1037/a0016539

McCutcheon, L. E. (1998). Stereotyping the nocturnal person: Findings with some alarming implications. Journal of Social Psychology, 138, 411– 413. doi:10.1080/00224549809600395

Moen, P., Kelly, E., & Hill, R. (2011). Does enhancing work–time control and flexibility reduce turnover? A naturally occurring experiment. Social Problems, 58, 69 –98. doi:10.1525/sp.2011.58.1.69

Piccolo, R., & Colquitt, J. (2006). Transformational leadership and job behaviors: The mediating role of core job characteristics. Academy of Management Journal, 49, 327–340. doi:10.2307/20159766

Podsakoff, P., & MacKenzie, S. (1989). A second-generation measure of organizational citizenship behavior. Unpublished manuscript, Indiana University, Bloomington, Department of Management and Entrepre- neurship.

Reb, J., & Cropanzano, R. (2007). Evaluating dynamic performance: The influence of salient gestalt characteristics on performance ratings. Jour- nal of Applied Psychology, 92, 490 – 499. doi:10.1037/0021-9010.92.2 .490

Roberts, B., Bogg, T., Walton, K., Chernyshenko, O., & Stark, S. (2004). A lexical investigation of the lower order structure of conscientiousness. Journal of Research in Personality, 38, 164 –178. doi:10.1016/S0092- 6566(03)00065-5

Roenneberg, T., Kuehnle, T., Pramstaller, P., Ricken, J., Havel, M., Guth, A., & Merrow, M. (2004). A marker for the end of adolescence. Current Biology, 14, R1038 –R1039. doi:10.1016/j.cub.2004.11.039

Rotundo, M., & Sackett, P. (2002). The relative importance of task, citizenship, and counterproductive performance to global ratings of job performance: A policy-capturing approach. Journal of Applied Psychol- ogy, 87, 66 – 80. doi:10.1037/0021-9010.87.1.66

Salgado, J. (1997). The Five Factor Model of personality and job perfor- mance in European community. Journal of Applied Psychology, 82, 30 – 43. doi:10.1037/0021-9010.82.1.30

Scott, K., & Brown, D. (2006). Female first, leader second? Gender bias in the encoding of leadership behavior. Organizational Behavior and Hu- man Decision Processes, 101, 230 –242. doi:10.1016/j.obhdp.2006.06 .002

Smith, C., Folkard, S., Schmieder, R., Parra, L., Spelten, E., Almiral, H., . . . Tisak, J. (2002). Investigation of morning– evening orienta- tion in six countries using the preferences scale. Personality and Individual Differences, 32, 949 –968. doi:10.1016/S0191- 8869(01)00098-8

Tajfel, H. E. (1978). Differentiation between social groups: Studies in the social psychology of intergroup relations. Oxford, England: Academic Press.

Tajfel, H., & Turner, J. (1986). The social identity theory of inter-group behavior. In S. Worchel & L. W. Austin (Eds.), Psychology of inter- group relations (pp. 7–24). Chicago, IL: Nelson-Hall.

Terry, D., & Hogg, M. (1996). Group norms and the attitude– behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22, 776 –793. doi:10.1177/0146167296228002

Tett, R., Jackson, D., & Rothstein, M. (1991). Personality measures as predictors of job performance: A meta-analytic review. Personnel Psy- chology, 44, 703–742. doi:10.1111/j.1744-6570.1991.tb00696.x

Tsaousis, I. (2010). Circadian preferences and personality traits: A meta- analysis. European Journal of Personality, 24, 356 –373. doi:10.1002/ per.754

Uhlmann, E., Leavitt, K., Menges, J., Koopman, J., Howe, M., & Johnson, R. (2012). Getting explicit about the implicit: A taxonomy of implicit measures and guide for their use in organizational research. Organizational Research Methods, 15, 553– 601. doi:10.1177/10944 28112442750

van Dam, K. (2003). Trait perception in the employment interview: A Five-Factor Model perspective. International Journal of Selection and Assessment, 11, 43–55. doi:10.1111/1468-2389.00225

(Appendix follows)

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1297PERCEPTION AND MORNING EMPLOYEES

Appendix

Study 3 Experimental Manipulation

Early Start Time Condition Digital Analysis Corporation (DAC) is a business consulting firm based in Bothell, Washington. It is generally perceived to be a good

place to work. John is a recent hire at DAC. Below is a graph of John’s recent objective performance report, which is based on the number of client contracts he fulfills.

One of the perks of working at DAC is its flexible work scheduling program. DAC allows its employees to arrive at the office anytime between 7 a.m. and 11 a.m. and to leave anytime between 3 p.m. and 7 p.m, as long as they work in 8-hr shifts. Generally, he chooses to set an early schedule. Most days, he comes in at 7 a.m. and leaves at 3 p.m.

-6,000

-5,000

-4,000

-3,000

-2,000

-1,000

0

1,000

2,000

3,000

4,000

5,000

6,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Week

P er fo rm

an ce

(Appendix continues)

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1298 YAM, FEHR, AND BARNES

Late Start Time Condition Digital Analysis Corporation (DAC) is a business consulting firm based in Bothell, Washington. It is generally perceived to be a good

place to work. John is a recent hire at DAC. Below is a graph of John’s recent objective performance report, which is based on the number of client contracts he fulfills.

One of the perks of working at DAC is its flexible work scheduling program. DAC allows its employees to arrive at the office anytime between 7 a.m. and 11 a.m. and to leave anytime between 3 p.m. and 7 p.m., as long as they work in 8-hr shifts. Generally, he chooses to set a late schedule. Most days, he comes in at 11 a.m. and leaves at 7 p.m.

-6,000

-5,000

-4,000

-3,000

-2,000

-1,000

0

1,000

2,000

3,000

4,000

5,000

6,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Week

P er fo rm an ce

Received July 31, 2013 Revision received March 21, 2014

Accepted May 6, 2014 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1299PERCEPTION AND MORNING EMPLOYEES