the thesis to end all thesis

54
AN EXPLORATION OF GROUP GENDER COMPOSITION: ATTRIBUTIONS MADE FOR SUCCESS AND FAILURE by Rebekah C. DeVore © 2015 Rebekah C. DeVore A thesis submitted in partial fulfillment of the requirements for

Upload: rebekah-devore

Post on 10-Jan-2017

55 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The Thesis to End all Thesis

AN EXPLORATION OF GROUP GENDER COMPOSITION:

ATTRIBUTIONS MADE FOR SUCCESS AND FAILURE

by

Rebekah C. DeVore

© 2015 Rebekah C. DeVore

A thesissubmitted in partial fulfillment

of the requirements for the degree of Master of Arts

Steinhardt School of Education, Culture, and Human Development Department of Applied Psychology

May 2015

Page 2: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 2

AN EXPLORATION OF GROUP GENDER COMPOSITION:

ATTRIBUTIONS MADE FOR SUCCESS AND FAILURE

by

Rebekah C. DeVore

__________________ Date___________ Research Mentor

__________________ Date___________ Program Director

Page 3: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 3

Acknowledgements

First and foremost, I am grateful to my family for their consistent love, support, and

encouragement, their belief in me helps me to achieve more than I thought possible. I

would like to express my sincere appreciation to Lindy Gullet and Tessa West for sharing

their expertise, guidance and encouragement each step a long the way. I am also thankful

to the HDSI Department Faculty for their help and support. I deeply appreciate Nathan

for his consistent encouragement at times when I felt I wanted to give up. My gratitude

extends to all who directly or indirectly helped me accomplish this venture. Without the

help of many this would not have been possible, and for their support I am grateful.

Page 4: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 4

Table of Contents

Acknowledgements……………………………………………………...3

List of Illustrations……………………………………………………...5

Abstract………………………………………………………………….6

Literature Review……………………………………………………….7

Methods………………………………………………………………….13

Results…………………………………………………………………...18

Discussion………………………………………………………………..21

References……………………………………………………………….27

Appendices………………………………………………………………32

Page 5: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 5

List of Illustrations

1. Performance Attribution by Gender……………………………….322. Example Vehicle Grid……………………………………………..333. Example Waiting Room…………………………………………...344. Success Manipulation…...…………………………………………355. Failure Manipulation………………………………………………35

Page 6: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 6

Abstract

Team-based work is becoming increasingly popular in the workplace. Recent research

shows that members of groups with more females than males perceive their group as less

competent and effective, regardless of their performance ability. The present research

examines how a group’s gender composition affects group members’ attribution

evaluations (i.e. explanations of whether the group’s performance was due to effort, luck,

skill, or task difficulty) and domain specific self-esteem (i.e. how participants feel about

themselves within the group). Participants completed a male-typed group task in a virtual

setting. Groups each included five members. Participants were randomly assigned to a

group of five that included either three females (female dominant) or two females (male

dominant). At the end of the task, participants were told that their group either succeeded

(performed better than average) or failed (performed below average) at the task, and

evaluated their group’s performance. Participants in the success condition evaluated their

group as performing significantly better than participants in the fail condition. Results

suggest that group gender composition affects how people use effort to explain

performance. However, the interaction between success versus failure and group gender

composition was not significant for any of the attributions (i.e. effort, luck, skill or task-

difficulty). The results are inconsistent with previous research on attribution and no

significant effects were found based on the predicted hypotheses. Working in groups is a

vital part of most careers; implications of gender-biased attributions in the workplace are

discussed.

Keywords: gender, attributions, group-gender, gender bias, gender discrimination, group-

gender composition.

Page 7: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 7

An Exploration of Group Gender Composition:

Attributions Made for Success and Failure

This research explores how group gender composition affects attribution

evaluations (i.e. explanations of whether the group’s performance was due to effort, luck,

skill, or task difficulty) and domain-specific self-esteem (i.e. how participants feel about

themselves within the group). Gender discrimination is a persistent challenge in the

workplace (Corbett & Hill, 2012; CAWP, 2014; Catalyst, 2013; Office of Science and

Technology Policy, 2014), and there is a substantial lack of women in roles that are

considered to be traditionally male-typed, (e.g. engineering or math related roles) (Lyness

2002, Powell 1999). Attributions made for success and failure (i.e. inferences made about

why people succeeded or failed) play an important role in future expectations and

motivation (Weiner, 1985, 2001, 2013). Attributions are evaluations that attempt to

explain the cause of an event. The four causal attributions most widely recognized are

effort, luck, ability, and task difficulty (Weiner, 1985, 2001, 2013). There are certain

gender biases found in the attribution process when evaluating success or failure. For

instance, people tend to explain why a woman succeeds differently than how they explain

why a man succeeds (Deaux, 1987; Dickhäuser and Meyer 2006; Jacobs and Eccles

1992; Swinton et al. 2011).

For example, if a woman is successful in math, people are more likely to explain

her success as being due to her effort; if a man is successful in math, people are more

likely to explain his success as being due to his innate ability. However, researchers have

not yet explored whether group gender composition, rather than a target’s gender, can

affect attributions made about a task groups’ success or failure. Therefore, the present

Page 8: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 8

work examines how the gender composition of a group influences group members’

explanations of their group’s success or failure (e.g. my group success was due to ability;

my groups failure was due to effort), and the way members feel about themselves in the

group (e.g. I am unable to do my best in this group).

Attribution

Attribution theory has been a dominant concept in social and educational

psychology for the past three decades. The present research pulls from the causal

attribution model based on Bernard Weiner’s approach to attribution achievement and

motivation (1985, 2001, 2013). This theory assumes that people infer causes of behavior

and events (e.g. when a student fails a test they ascribe a reason for why they believe that

happened). According to the theory, there are three causal dimensions: locus of causality

(internal vs. external), stability (stable vs. unstable), and controllability (control vs. no

control).

Expectancy

Research specific to the effect of gender and attributions shows that gender tends

to affect how attributions are made on the dimension of stability. Because females are not

expected to do well on male-typed tasks, people tend to explain a woman’s success using

unstable attributions (i.e. effort or luck). In contrast, when a woman fails at a male-typed

task, people tend to explain her failure using stable causes. Because males are expected to

perform well on male-typed tasks, the reverse is true; when a man fails, people tend to

explain his failure using unstable causes (i.e. effort or luck). When he succeeds, people

will tend to explain his success using stable causes (i.e. ability). Deaux (1984, 1987)

explains this phenomenon by using the expectancy model; when performance is

Page 9: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 9

consistent with expectations, outcomes will be attributed to stable causes (ability or task

difficulty) and when they are inconsistent with expectations, they will be attributed to

causes that are unstable (effort or luck). Because careers related to math and engineering

have been traditionally male-typed, it is unexpected for women to succeed and excel in

them. Based on Weiner’s theory of attribution and Deaux’s expectancy model,

expectations for women to be unsuccessful in male-typed careers influences attribution

and future motivation, making it more difficult for women to succeed and excel in male-

typed careers even though they are able.

In an experiment, Deaux and Emswiller (1974) had males and females observe

and evaluate a confederate’s performance on a task. The task was a male-typed memory

recall task. The confederates were comprised of both males and females, all confederates

performed above average on every task. When participants were evaluating male

confederates on the recall task they tended to report that his success was due to ability

more so than luck. When both female and male participants were evaluating a woman

confederate on the same task, (who performed equally as well), participants rated her

success as being more based on luck.

Similar findings occur in educational fields of research. Math in educational

settings, for example, is recognized as male-typed. Even though there are not gender gaps

in math achievement, males tend to dominate occupations that are math related (Eccles

2007, Halpern et al. 2007; Hyde and Mertz 2009; Scafidi and Bui 2010). Educational

research, pertaining to attributions and gender bias, consistently show the tendency for

boys success in math to be attributed to ability and girls success to effort, while boys

failure is attributed to a lack of effort and girls failure to a lack of ability. Gender biased

Page 10: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 10

attributions have been demonstrated in a variety of empirical studies among parents

(Jacobs and Eccles 1992; Yee and Eccles 1988), students (Bornhold and Möller 2003;

Dickhäuser and Meyer 2006; Jacobs and Eccles 1992; Ryckman and Peckham 1987;

Stipek and Gralinkski 1991; Swinton et al. 2011), and teachers (Fennema et al. 1990;

Räty et al. 2002; Tiedemann 2002). Based on research in both social and educational

psychology, attributions made for an individual’s success and failure vary based on their

gender (see Appendix A for table of attribution by gender).

Group Gender Composition

Although research on attributions has only examined success and failure in

settings where men and women are evaluated independently (i.e. not as part of a group),

there exists a small body of research that suggests how attributions of success and failure

may occur within group settings. West et al. (2012) discovered that when people are in

gender diverse groups, the more women there are, the less competent the group is rated.

Five-person task groups worked together to complete a male-typed task. Groups varied in

gender composition; they either had two, three, or four women. The male-typed task was

to build a replica of a complex model made of Legos™. All members had to work

together to complete the task. When groups completed the task they brought their replica

to a judge. The judge would reject the replica if it did not match the model and the group

would have to figure out what needed to change. They would bring the replica to the

judge until it matched the model correctly. After the replica was accepted participants

evaluated the task contributions of other group members, and rated the overall

effectiveness of their group. Participant’s evaluated group members’ task contributions

and the overall effectiveness of their group. Ten weeks after the study, participants were

Page 11: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 11

asked to report how interested they were in working with their group again on a graded

project. Measures of task contributions, overall group effectiveness, and desire to work

with the group in the future were reported on 7-point likert scales, 1 (disagree strongly) to

7 (agree strongly). The number of times a model was rejected was used to measure

objective performance. The extent to which an individual group member contributed to,

was focused on, was competent at, and helped the team complete the task was used to

measure the construct of task-contribution. The measure of group effectiveness included

three questions: the team as a whole worked well together, my team was good at

coordinating the work of all members, and I would like to work with the team in the

future. Results showed that as the number of women in the group increased the

judgments of the members’ task contributions, and the overall group effectiveness ratings

decreased, and when asked 10 weeks later, groups with more women had less desire to

work with their group again. The findings show that gender did not predict performance

on the task, but it did predict people’s opinions of their groups.

Although West et al. (2012) did consider evaluations made about individual group

members’ task contributions, the groups’ overall effectiveness, and desire to work with

the group in the future; explanations made for success or failure were not considered.

Some of the individual items used in measurement align with effort and others align with

skill. For example, within the task contribution construct, the extent to which a member

contributed to, was focused on, or helped complete the task would align with an effort

attribution; the question asking about a member’s competence would align with an ability

attribution. As such, the current research will consider the different dimensions on which

an individual can be evaluated. In addition, the current research will investigate how the

Page 12: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 12

group’s actual performance moderates the relationship between group gender

composition and attributions about group members. I hypothesized that when evaluating

the group as a whole, members of a female dominant group would attribute their group’s

success to unstable causes (effort or luck) and failure to stable causes (ability or difficulty

of task), and the opposite would be found in male dominant groups; group’s success

would be attributed to stable causes and failure to unstable causes.

Domain Specific Self-Esteem

The way group gender composition interacts with success and failure attributions

may affect domain specific self-esteem. Research reveals that self-esteem can fluctuate

based on the context a person is in (Kernis, 2005), and that self-esteem measures are

sensitive to context (Bosson, 2000). A considerable amount of work has been directed

towards understanding self-esteem within organizational contexts. A meta-analysis of

organization-based self-esteem found: the way a person feels within an organization

significantly affects job satisfaction, commitment to the organization, motivation,

performance in-role, intentions of turnover, and other organization related behaviors and

attitudes (Pierce, 2004). There is also a considerable amount of work showing that

domain specific self-esteem varies significantly depending on gender (Gentile, 2009).

Because group work is becoming increasingly popular in the workplace and organization

based self-esteem predicts behavior and attitude outcomes, looking at how participants

feel in a group is important. In this paper, domain specific self-esteem is used to define

how the participant feels about being a member of their group, this is specific to the

domain because it is not attempting to measure how the person feels about themselves

overall (global self-esteem). If biases occur based on gender composition, success versus

Page 13: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 13

failure, and the attributions made, it will be interesting to see if and how domain-specific

self-esteem is affected.

Based on the previous research I predicted that when evaluating the group as a

whole, members of a female dominant group would attribute their group’s success to

unstable causes (effort or luck) and failure to stable causes (ability or difficulty of task),

and the opposite would be found in male dominant groups; group’s success would be

attributed to stable causes and failure to unstable causes. A second prediction was that the

interaction between the condition of success or failure and the attribution made would

predict domain-specific self-esteem (e.g. if the group is female dominant and success is

attributed to unstable causes the participant would report a relatively lower domain

specific self-esteem).

Methods

Participants

Participants were 366 (189 female) Mechanical Turk users (Mage= 33.54, age

range: 18-70 years; 67% white, 23% Asian, 6% Latino, 6% Black, 2% Native

American/Alaskan Native, 1% multiracial, and 1% reported other). Participants were

Mechanical Turk users who chose take part in the study. The study was made using

Qualtrics and uploaded to Mechanical Turk. Participants were compensated with $1.00 if

they completed the study. They were randomly assigned to a female dominant (three

females: two males), or a male dominant (three males: two females) group, and a success

or fail condition.

Page 14: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 14

Procedure

Participants worked in an “online group” attempting to obtain the fastest group

score while playing the game Rush Hour®. Groups all included five members; one was

the participant; the remaining four members were computer bots (i.e. computer-simulated

confederates) created for the study. The participant was led to believe the other members

of the group were other Mechanic Turk users who were taking the study. The computer-

simulated confederates were created so that all conditions were constant and controlled.

The two types of groups were male dominant (three males, two females) and female

dominant (three females, two males). The computer bots’ gender was manipulated such

that participants were either in a male dominant or female dominant group. The gender

was manipulated by using a female or male icon next to a gender specific name. The

names used for the computer bots were dependent upon the condition: Member 1 =

Michael/Emily; Member 2 = Matthew/Jessica; Member 3 = Jacob/Ashley; Member 4:

Participant; and Member 5 = Christopher/Sarah. Names were chosen based on the most

popular baby names in 1991 list, a year when a lot of potential participants were born. It

was programmed so that the groups’ dominant gender was randomized and dependent on

the participants’ specified gender, so that there was never only one female or one male in

the group.

After consent, participants were asked to give their first name and demographic

information. The participants were told they would be working on a task called Rush

Hour® in a group with four other members. They were informed the task would require

them to use their spatial reasoning to free vehicles from a traffic jam and told the five

participants with the best performance ratings would receive a $4 bonus. The participant

Page 15: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 15

was shown a square grid with cars and trucks in a randomized pattern (see Appendix B

for an example vehicle grid).

Participants were instructed that the goal of the game was to free the specified

vehicle from the traffic jam and through the exit as quickly as possible. Vehicles were

each assigned a number (one through five). Each member was assigned to a number to

specify the order in which they took their turn and the vehicles they were allowed to

move during their turn. The participant was always group member four, and the other

members names were altered based on the group condition. All comments and

performance strategies of computer bots were constant, except for the change of name

and icon (i.e. female or male). Each car or truck could only move horizontally or

vertically based on the positioning and had to be moved one at a time in order to get the

specified vehicle out of the traffic jam. They were instructed that they could only move

vehicles with their assigned number. Afterwards participants entered an “online” waiting

room and were led to believe others who joined the waiting room were also participants

who would be their group members. They were shown other group member’s names and

an image signifying their gender (see Appendix C for an example of the online waiting

room). Once all five members (confederate computer bots plus the participant) arrived in

the waiting room participants were automatically directed to the next page to begin the

task.

There were two rounds, each with a different grid and car to free. Both rounds

included a planning period and an action period. Participants were shown the grid at the

beginning of each round then given one minute and 30 seconds to type their strategy to

share with their group during the planning period. After time was up they were

Page 16: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 16

automatically directed to the next page. The page contained all of the group members’

strategies. They were given two minutes to read their group members strategy responses.

After two minutes they were automatically directed to the next page to begin the action

period. During this period the participant was only able to move their assigned vehicle(s),

when it was their turn to do so (e.g. Group member one was only able to move vehicles

labeled with the number one, and this member went first). Each member of the group had

their sequential turn to move their specified vehicle (participant was always fourth). The

participant saw each member’s board during that members turn, they also saw how long

it took members to complete the turn, what they decided to do, and if they wrote any

comments to the group. During the participants turn they were given 20 seconds to

choose how they wanted to move their specified vehicle(s) and an option to write a

comment to their group. Turns continued until the specified vehicle was freed through the

exit. After both rounds were completed the participant was moved into the evaluation

period. During this time, participants were shown an analysis of their group’s

performance. The times were manipulated so that if they were in the failure condition the

average time of other groups was below their group’s performance average, and in the

success condition the average time of other groups was above their groups performance

average. The group’s performance time average was held constant, and the average they

believed were other groups was manipulated for the failure or success condition.

Numbers used were based on how long the different computer bot members were

programmed to take during their respective turns plus the planning periods and the

estimated time it would take participants. For the success condition participants were

shown that the average time of the groups who had already completed the study was 9

Page 17: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 17

minutes and 24 seconds; their groups average time was shown as 8 minutes and 14

seconds; the individual group members times were shown as follows: Member One, 1

minute and 10 seconds; Member Two, 1 minute and 0 seconds; Member Three, 0 minutes

and 57 seconds; Member Four, 1 minute and 3 seconds; Member Five, 0 minutes and 50

seconds. Each group member’s number, name, and then completion time were shown on

the group’s analysis. For the failure condition all numbers were shown the same as the

success condition, the only number that changed was the average completion time of

other groups who had previously completed the study. The number shown for the average

of all the groups who had completed the study in the failure condition was 7 minutes and

10 seconds. See Appendix D for examples of each condition. Participants were expected

to believe that their group did relatively better or worse than the average performance of

other groups. Participants evaluated their group’s performance and member’s

performances individually. Throughout the evaluation period a transcript of the groups

analysis was available for viewing, so that participants had the option to use the analysis

to help them evaluate their group.

Measures

Groups were manipulated to either fail or succeed. The manipulation was based

on the groups timing. Groups that were manipulated to fail had a completion time that

was slower than the reported average of all that groups that had participated. Participants

in groups that were manipulated to succeed were shown a completion time that was faster

than the reported average. The times used were

Participants evaluated themselves, and the group as whole. The attribution measurement

consisted of four items about why they believed their group succeeded or failed

Page 18: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 18

(effort/luck vs. ability). The attribution measurement was based on success and failure

attribution questions used in previous research (Räty et al., 2002; Deaux & Emswiller,

1974). There were four attribution questions. The questions were each on a seven point

likert scale, ranging from not at all due to (skill, effort, luck, or task difficulty) to purely

due to (skill, effort, luck or task difficulty). The questions asked were: (a) to what extent

was your group’s performance due to skill; (b) to what extent was your group’s

performance due to effort; (c) to what extent was your group’s performance due to luck;

and (d) to what extent was your group’s performance due to the difficulty of the task?

Participants then answered four self-evaluation items about how they felt about

themselves in their group. Items were based on the Rosenberg Self-Esteem Scale and

altered to be specific to the group context (Rosenberg, 1987). The questions were on a

five-point likert scale ranging from strongly disagree to strongly agree. The questions

were: (1) In this group I was able to do my best; (2) I was able to share a number of my

good qualities with this group; (3) I feel I do not have much to be proud of in this group;

and (4) I have a positive attitude towards myself in this group. Question three was reverse

coded after the data was collected. To check the success versus failure manipulation, the

question: “In terms of performance, how well has your group done on the task?” was

asked. Findings should show a significant difference between the success and failure

condition. Participants should have felt that their group did not perform as well if they

were in the failure condition.

Results

Simple linear regressions were run on group gender, participant gender, success

versus failure, and the interactions between the variables. Regression analyses were run

Page 19: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 19

using SPSS statistical software. The dependent variables were effort, skill, luck, task-

difficulty, and domain-specific self-esteem. A manipulation check was run to ensure

participants were experiencing the success and failure conditions differently.

Effort

If a participant was in a female-dominant group and the group succeeded, a

trending effect showed participants were more likely to attribute that success to effort, but

this finding was not significant, t(358) = 1.558, ps > .120. Participants were significantly

more likely to attribute their group’s performance to effort if they were in the success

condition, t(358) = 3.704, p < .001. Participants reported their group's performance as

more due to effort if they were in a male rather than a female-dominant group, but this

effect was not significant, t(358) = -1.629, ps > .104.

Skill

Participants in a male-dominant group were not significantly more likely to

attribute their success to skill if they succeeded, t(358) = -.032, ps > .975. Participants did

not view their group’s performance as being more due to skill if they were in a male,

rather than a female-dominant group, regardless of the success or failure condition, t(358)

= -.083, ps > .934. The way participants reported their group’s performance as being due

to skill did not vary significantly between the success and failure conditions, t(358)

= .208, ps > .208.

Task-Difficulty

Participants in female-dominant groups that succeeded were not significantly less

likely to attribute their group’s performance to task difficulty than participants in male-

dominant groups who succeeded, t(358) = .392, ps > .695. Difficulty of task evaluations

Page 20: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 20

did not vary based on the group gender composition, t(358) = -1.144, ps > .253.

Participants did not vary in their task-difficulty attributions based on the success or

failure of the group, t(358) = .884, ps > .378.

Luck

Participants in female-dominant groups that succeeded were not significantly

more likely to attribute their group’s performance to luck, t(358) = .-.809, ps > .419. Luck

evaluations did not vary based on the group gender composition, t(358) = -.307, ps

> .759. Participants did not vary in their luck attributions based on the success or failure

of the group, t(358) = .237, ps > .813.

Domain-Specific Self-Esteem

Participant domain-specific self-esteem did not vary dependent on the interaction

between group gender composition and the success versus failure condition, t(358) =

-.434, ps > .665. There was no significant effect of group gender composition on domain

specific self-esteem, t(358) = -.265, ps > .792. Participants in the fail condition reported

significantly lower domain specific self-esteem, t(358) = 2.330, p < .05

Overview

There were no significance effects on attribution evaluations of luck, ability, or

task difficulty. Participants reported their group's performance as being more due to effort

when they were in male-dominant rather than female-dominant groups, but the effect was

not significant. Participants were more likely to attribute success to effort in a female

dominant group, but the effect was not significant. Participants were significantly more

likely to attribute success to effort when they were in the failure condition. Being in a

female or male dominant group did not have a significant effect on domain-specific self-

Page 21: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 21

esteem. Participants reported significantly lower domain-specific self-esteem if they were

in a group that failed. A manipulation check revealed participants did feel differently

about their team’s performance based on the success or failure condition. They felt their

team did significantly worse if they were in the fail condition, t(358) = 4.76, p < .001.

This was expected, and demonstrates the manipulation was effective.

Discussion

Although there were significant effects found in the results, there was no evidence

in this study to support the central hypotheses; female dominant groups would be more

likely to attribute their group’s success to unstable factors (effort or luck/chance) and

their group’s failure to stable factors (ability or task difficulty), and male groups would

be more likely to attribute their group’s success to stable factors and failure to unstable

factors. Among the differing performance attribution outcomes, evaluations based on

effort showed the most movement within the data. Trending effects revealed that overall

male-dominant group performance was more due to effort; in the success condition

female-dominant group performance was more likely to be attributed to effort than male

dominant groups. Being in a female group and attributing success to effort is in line with

previous research and the main hypotheses, because effort is an unstable attribution, and

it is unexpected that females should succeed at a male-typed task, but this effect was only

trending and not significant. The trend showing that regardless of success or failure male-

dominant groups’ performance is more likely due to effort is unexpected and not

supported by previous attribution research. Pertaining to effort there was a significant

effect showing that participants were more likely to attribute performance to effort if they

were in the success condition. Performance attribution results suggest that effort

Page 22: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 22

attributions are used for failure in a male-dominant group, more so than in female-

dominant groups who fail. Because effort was significantly more likely to be attributed in

successful groups, trending in female-dominant groups who succeeded, and in male-

dominant groups for both success and failure. Attributing effort to male-dominant groups

who failed is in line with the main hypotheses because effort is an unstable attribution,

and it is unexpected that males would fail at a male-typed task.

Even though the results for skill, task-difficulty and luck were not significant,

they are still informative. It is possible that those concepts should be captured in a

different way. The results suggest that people in groups use effort to explain success and

failure more so than they use any other attribution explanation (i.e. skill, luck, or task

difficulty). People in groups use effort to explain both success and failure, if there are

more males in a group it is more likely that both success and failure will be attributed to

effort. This is helpful because if the group fails they are able to just blame it on not

putting in enough effort. If there are more females in the group and the group succeeds it

is likely that it will be attributed to effort, but if there are more females in the group and

the group fails it is less likely that the performance will be attributed to effort. It is

inconclusive how groups with more females explain their failure. It is unfortunate that

groups with more females do not explain failure as due to lack of effort, because using

effort to explain failure makes it more probable for future motivation to achieve at the

task.

The effect of group gender composition and attribution on domain specific self-

esteem was not explored because there was no main effect found in the interaction

between gender composition and attributions for success or failure. It is still possible that

Page 23: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 23

domain specific self-esteem is affected by differing attributions for group success or

failure, but this study was unable to capture an interaction between group gender

composition and attributions for success or failure. Domain specific self-esteem was

reported as significantly less in groups that failed. This supports previous research

claiming that the context influences self-esteem. This result suggests future research

could explore workplace interventions based on resilient responses to lowered self-

esteem due to group failure. Previous research shows that employees who feel better in

their work environment produce more; interventions aimed at learning from failure

instead of becoming defeated by it could help alleviate turnover and under-productive

employees.

Limitations

The study had limitations that should be discussed. Because there is extensive

research supporting individual attribution gender bias; when women perform a male-

typed task, attributions made for their successes are more likely to be unstable and males

successes more likely to be stable; it is possible that participants did not view the task in

this study as male-typed. Therefore, the attributions made for success or failure would not

have any predicted effects.

In previous attribution research there are considerable inconsistencies found in

measurement. Studies use a variety of measures to capture attribution effects. There are

measure that ask direct questions about each attribution and measures that ask indirectly

to capture the idea of each attribution. There are measures that use likert scales for each

individual attribution and bipolar likert scales with the attributions paired against one

another (e.g. skill on one end of the scale and ability on the other). There are also

Page 24: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 24

measures that ask open-ended questions and code the answers to gauge which attribution

it was most in line with. The lack of a valid, reliable measure used across many studies is

a critique of the theory. It is possible that the lack of an accurate measure has made it

difficult to capture the effect of group gender composition on attributions in this study.

Wording the questions as “how much was your groups performance due to

effort…” (for both the success and failure conditions) makes the question symmetrical,

but it does not accurately reflect the concept of the attribution. If, on the failure condition

the question was formatted as “how much was your group’s performance due to lack of

effort,” it more accurately describes the concept but cannot be used as the same question

in the success condition. This presents a conceptual problem; some previous research

uses a bi-polar scale with effort at one end and ability or skill on the other. This more

accurately captures the theoretical concepts but is not as accurate of a measurement. If

participants rate a high score on effort they are automatically rating a low score on effort,

capturing an effect that may not be accurate. These conceptual problems in previous

research make it difficult to identify and explain attribution bias outcomes and processes.

Participants in this study evaluated their group mates with a very limited amount

of information. During evaluation, participants were shown only a numerical datum about

each group member. Having a transcript of each member’s comments throughout the

study available to view during the evaluation period could have made a difference in the

participant’s evaluations of the group. Evaluation based on viewing members’ comments

would be closer to real life; people evaluate group members based on comments made

and not just data output.

Page 25: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 25

Having the sample limited to those who are working in male-typed careers might

reveal different trends. This study did not ask about the participants’ career fields. The

possible effects of this study are directed towards those working in traditionally male-

typed careers, limiting the sample to those who are already in this field may prove more

productive.

Future Research

Research exploring group gender composition and its effect on success and failure

attributions are influential in understanding how to eliminate gender bias in the

workplace. Because there is extensive research exploring and defining the processes of

gender bias on an individual level, and the context is instrumental in producing different

effects; the context of being in a group with differing amounts of males and females

could have processes unique to being in a group.

Future research could explore, more in depth, how effort attributions are affected

by the interaction between group success versus failure and gender composition. The

interaction between individual gender, group gender, and company gender (proportion of

males to females in the entire company) and how success or failure on all three levels

(individual, group, company) changes attributions or motivation within individuals,

would be interesting to explore. If gender biased attributions differ dependent on the

overall companies’ gender, the task groups’ gender, and the individual gender, it would

be helpful to know why and how so future research would be able to develop productive

interventions aimed at changing systemic gender bias.

Qualitative research about women in traditionally male-typed careers that

explores how women and men feel when task groups succeed or fail could provide deeper

Page 26: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 26

insight and understanding in the processes that occur within group work. Having in-depth

interviews with men and women before, during, and after they have worked within task-

groups that have either succeeded or failed, could reveal processes and trends that were

not suggested in this paper. It would also provide understanding into the ways attributions

change over time and if they vary based on individual and group gender.

If a group’s gender composition has different effects and processes than merely

the summation of the individual gender effects and processes, then the way we study and

attempt to change gender bias in the workplace should be adjusted accordingly. Group

work is an inescapable part of most workplaces and women are still having a difficult

time moving up the ranks within traditionally male-typed careers (even though research

shows women to be equally capable). Therefore, focusing attention on how groups think

about success or failure based on group gender could reveal new conclusions;

conclusions necessary to relieve unfounded bias and allowing women to be successful in

traditionally male-typed careers because of their ability.

Page 27: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 27

References

Bornholt, L., & Möller, J. (2003). Attributions about achievement and intentions about

further study in social context. Social Psychology of Education, 6, 217–231.

Bosson, J. K., Swann Jr, W. B., & Pennebaker, J. W. (2000). Stalking the perfect measure

of implicit self-esteem: The blind men and the elephant revisited?. Journal of

personality and social psychology, 79(4), 631. doi:10.1037/0022-3514.79.4.631

CAWP Center for American Women and Politics (2014). Women in the U.S. congress

2014. Eagle Institute of Politics, Rutgers University. Retrieved from

http://www.cawp.rutgers.edu/fast_facts/levels_of_office/documents/cong.pdf.

Catalyst (2013). Statistical overview of women in the workplace. Knowledge Center.

Retrived from http://www.catalyst.org/knowledge/statistical-overview-women

workplace.

Chapman, K. J., Meuter, M., Toy, D., & Wright, L. (2006). Can’t we pick our own

groups? The influence of group selection method on group dynamics and

outcomes. Journal of Management Education, 30(4), 557-569.

Corbett, C. & Hill, C., (2012). Graduating to a pay gap: The earnings of men and

women one year after college. Washington, DC: AAUW American Association of

University Women.

Deaux, K. (1984). From individual differences to social categories: Analysis of a decade's

research on gender. American Psychologist, 39(2),105.

Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of

gender related behavior. Psychological review, 94(3), 369.

Deaux, K., & Emsweiler, T. (1974). Explanations of successful performance of sex

Page 28: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 28

linked tasks: What is skill for the male is luck for the female. Journal of

Personality and Social Psychology, 29, 80-85.

Dickhäuser, O., & Meyer, W. (2006). Gender differences in young children’s math ability

attributions. Psychology Science, 48(1), 3–16.

Eccles, J. S. (2007). Where are all the women? Gender differences in participation in

physical science and engineering. In S. J. Ceci, S. J. Ceci, W. M. Williams, & W.

M. Williams (Eds.), Why aren’t more women in science: Top researchers debate

the evidence (pp.199–210). Washington, DC: American Psychological

Association.

Fennema, E., Peterson, P. L., Carpenter, T. P., & Lubinski, C. A. (1990). Teachers’

attributions and beliefs about girls, boys, and mathematics. Educational Studies in

Mathematics, 21, 55–65.

Gentile, B., Grabe, S., Dolan-Pascoe, B., Twenge, J. M., Wells, B. E., & Maitino, A.

(2009). Gender differences in domain-specific self-esteem: A meta-

analysis. Review of General Psychology, 13(1), 34. doi:10.1037/a0013689

Halpern, D. F., Benbow, C. P., Geary, D. C., Gur, R. C., Hyde, J. S., & Gernsbacher, M.

A. (2007). The science of sex differences in science andmathematics.

Psychological Science in the Public Interest, 8(1), 1–51.

Heilman, M. E., & Haynes, M. C. (2005). No credit where credit is due: attributional

rationalization of women's success in male-female teams. Journal of Applied

Psychology, 90(5), 905.

Hyde, J. S., & Mertz, J. E. (2009). Gender, culture, and mathematics performance.

Proceedings of the National Academy of Sciences, 106(22), 8801–8807.

Page 29: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 29

Jacobs, J. E., & Eccles, J. S. (1992). The impact of mothers’ gender-role stereotypic

beliefs on mothers’ and children’s ability perceptions. Journal of Personality and

Social Psychology, 63, 932–944.

Kernis, M. H. (2005). Measuring self‐esteem in context: The importance of stability of

self esteem in psychological functioning. Journal of personality,73(6), 1569-

1605. doi: 10.1111/j.1467-6494.2005.00359.x

Lyness, K. S. (2002). Finding the key to the executive suite: Challenges for women and

people of color. In R. Silzer (Ed.), The 21st century executive: Innovative

practices for building leadership at the top (pp. 229–273). San Francisco: Jossey-

Bass.

Office of Science and Technology Policy (2014). Women in STEM: the White House.

Retrieved from http://www.whitehouse.gov/sites/default/files/microsites

/ostp/stem_factsheet_2013_07232013.pdf

Pierce, J. L., & Gardner, D. G. (2004). Self-esteem within the work and organizational

context: A review of the organization-based self-esteem literature. Journal of

Management, 30(5), 591-622. doi: 10.1016/j.jm.2003.10.001

Powell, G. N. (1999). Reflections on the glass ceiling: Recent trends and future prospects.

In G.N. Powell (Ed.), Handbook of gender and work (pp. 325–345). Thousand

Oaks, CA: Sage.

Räty, H., Vänskä, J., Kasanen, K.,&Kärkkäinen,R. (2002). Parents’ explanation of their

child’s Performance in mathematics and reading: A replication and extension of

Yee and Eccles. Sex Roles, 46, 121–128.

Page 30: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 30

Ryckman, D. B., & Peckham, P. D. (1987). Gender differences in attributions for success

and failure. The Journal of Early Adolescence, 7(1), 47–63.

Scafidi, T., & Bui, K. (2010). Gender similarities in math performance from middle

school through high school. Journal of Instructional Psychology, 37(3), 252–255.

Stipek, D. J., & Gralinski, J. H. (1991). Gender differences in children’s achievement

related beliefs and emotional responses to success and failure in mathematics.

Journal of Educational Psychology, 83, 361–371.

Swim, J.K., & Sanna, L. (1996). He’s skilled, she’s lucky: A meta-analysis of observers

attributions for women’s and men’s successes and failures. Personality and Social

Psychology Bulletin, 22, 507-519.

Swinton, A. D., Kurtz-Costes, B., Rowley, S. J., &Okeke-Adeyanju, N. (2011). A

longitudinal examination of African American adolescents. Child Development,

82(5), 1486–1500.

Takeda, S., & Homberg, F. (2014). The effects of gender on group work process and

achievement: an analysis through self‐and peer‐assessment. British Educational

Research Journal, 40(2), 373-396.

Tiedemann, J. (2002). Teachers’ gender stereotypes as determinants of teacher

perceptions in elementary school mathematics. Educational Studies in

Mathematics, 50, 49–62.

Weiner, B. (1985). An attributional theory of achievement motivation and emotion.

Psychological review, 92(4), 548.

Weiner, B. (2001). Intrapersonal and interpersonal theories of motivation from an

attribution perspective. In Student Motivation (pp. 17-30). Springer US.

Page 31: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 31

Weiner, B. (2013). Human motivation. Psychology Press.

West, T. V., Heilman, M. E., Gullett, L., Moss-Racusin, C. A., & Magee, J. C. (2012).

Building blocks of bias: Gender composition predicts male and female group

members’ evaluations of each other and the group. Journal of Experimental

Social Psychology, 48(5), 1209-1212.

Yee, D., & Eccles, J. S. (1988). Parent perceptions and attributions for children’s math

achievement. Sex Roles, 19, 317–333.

Page 32: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 32

Appendix APerformance Attribution by Gender

Page 33: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 33

Appendix BExample Vehicle Grid

Page 34: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 34

Appendix CExample Waiting Room

Page 35: The Thesis to End all Thesis

ATTRIBUTIONS FOR SUCCESS AND FAILURE IN GROUPS 35

Appendix DSuccess and Failure Manipulations

Figure 1. Success Manipulation. This figure is an image that illustrates what a participant saw if they were in the success condition.

Figure 2. Failure Manipulation. This figure is an image that illustrates what a participant saw if they were in the fail condition.