1. The Knowledge, Skill, and Ability Requirements for Teamwork
Stevens and Campion (1994)

2. knowledge, skill, and ability (KSA)2
knowledge, skill, and ability (KSA)
Furthermore, a focus on KSAs emphasizes attributes
which management can influence (e.g., via selection procedures or training
programs), rather than trait or dispositional attributes which are presumed to
be more stable characteristics of individuals that cannot be as readily influenced.
Focus on team rather than technical KSAs. The study also focuses on
The focus is on:
(1) KSAs rather than personality traits;
the history of success has been much greater with ability-based selection strategies.
KSAs emphasizes attributes which management can influence.
(2) team rather than technical KSAs; and
social and interpersonal requirements.
(3) the individual rather than team level of analysis
it often synthesizes the literature and infers the individual-level KSAs from the group- and organizational-level theories and findings.

3. Focus on the individual in the team.
the framework included only those attributes which are at the individual team member level of analysis, as opposed to the group or organizational levels.
That is, the focus was on how to hire individuals for teams and not, for example, on how best to configure some optimal combination of members for a team.

4. As each body of
literature is reviewed, the relevant theories and findings are examined for
underlying similarities, and then condensed and classified into a taxonomy. The
result is 2 major categories of KSAs, with 5 subcategories and 14 specific KSAs
as summarized in Table 1.
Work teams, therefore, appear to require employees who possess the
following three sub-categories of interpersonal KSAs: (1) Conflict Resolution
KSAs (i.e., the ability to recognize and encourage desirable, but discourage undesirable
team conflict; to recognize the type and source of conflict and implement
an appropriate resolution strategy; and to use integrative, rather than distributive,
approaches to negotiation); (2) Collaborative Problem Solving KSAs (i.e., the
ability to match the proper degree of participation to the problem; and to recognize
obstacles to collaborative problem solving and implement appropriate corrective
actions); and (3) Communication KSAs (i.e., the ability to recognize and
utilize decentralized networks to enhance communication; to communicate openly
and supportively; to listen nonevaluatively and use active listening techniques; to
match one's own nonverbal and verbal messages and to recognize and interpret
the nonverbal messages of others; and to engage in small talk and ritual greetings).
teams. Work teams, therefore, appear to require employees who possess the 4

5. 5
three subcategories of interpersonal KSAs are identified which
individual team members should possess in order to be effective team
contributors: (1) conflict resolution KSAs; (2) collaborative problem solving
KSAs; and (3) communication KSAs.
Interpersonal KSAs
According to Dyer (1984) we have very little systematic knowledge about which interpersonal skills are most desirable.
A goal of this study is to specify this domain of interpersonal team member capabilities in a way that will allow for pragmatic and meaningful operationalizations.

6. Interpersonal KSAs Conflict Resolution KSAs6
Interpersonal KSAs
Conflict Resolution KSAs
R1. The KSA to recognize and encourage desirable, but discourage undesirable, team conflict.
conflict can have negative or positive effects, depending upon its nature and amount, and how it is addressed
Conflict is unproductive when disagreements reach an impasse and incapacitate a team,
but constructive conflict allows teams to identify problems, develop solutions
The objective should be to manage conflict to achieve optimal team performance.

7. Interpersonal KSAs Conflict Resolution KSAs7
Team members should be able to recognize the type of conflict and
match it to the appropriate resolution strategy. For example, when conflict
arises from a miscommunication, it can typically be resolved through
questioning and listening techniques. When conflict arises from situational
factors, it can often be resolved through rearranging the situation, such as when
conflict over a distasteful task can be resolved by a rotating assignment schedule
(Deutsch, 1973; Holmes & Miller, 1976; Thomas, 1977).
Interpersonal KSAs
Conflict Resolution KSAs
R2. The KSA to recognize the type and source of conflict confronting the team and to implement an appropriate conflict resolution strategy.
Conflict can take many forms or emanate from many sources, including:
simple misunderstandings or miscommunications,
structural or situational constraints,
incompatible performance goals or rewards,
requirements for joint decision making,
differences in values, orientations, and objectives,
or even such seemingly innocuous elements as physical design of a work area.

8. As each body of
literature is reviewed, the relevant theories and findings are examined for
underlying similarities, and then condensed and classified into a taxonomy. The
result is 2 major categories of KSAs, with 5 subcategories and 14 specific KSAs
as summarized in Table 1.
Work teams, therefore, appear to require employees who possess the
following three sub-categories of interpersonal KSAs: (1) Conflict Resolution
KSAs (i.e., the ability to recognize and encourage desirable, but discourage undesirable
team conflict; to recognize the type and source of conflict and implement
an appropriate resolution strategy; and to use integrative, rather than distributive,
approaches to negotiation); (2) Collaborative Problem Solving KSAs (i.e., the
ability to match the proper degree of participation to the problem; and to recognize
obstacles to collaborative problem solving and implement appropriate corrective
actions); and (3) Communication KSAs (i.e., the ability to recognize and
utilize decentralized networks to enhance communication; to communicate openly
and supportively; to listen nonevaluatively and use active listening techniques; to
match one's own nonverbal and verbal messages and to recognize and interpret
the nonverbal messages of others; and to engage in small talk and ritual greetings).
teams. Work teams, therefore, appear to require employees who possess the 8

9. Interpersonal KSAs Collaborative Problem Solving KSAs9
Participation by all team members in every decision may not always be
wise, however. The degree of participation should vary as a function of the
characteristics of the decision, such as the significance of a wrong decision,
importance of team acceptance, simplicity of the decision, availability of
information, and capability of the members (Vroom & Jago, 1978). Potential
drawbacks of participation include increased time requirements, lower quality
decisions if members lack the ability to contribute, diminished individual
responsibility, and the creation of an attitude of entitlement to involvement in
all decisions (Yukl, 1981).
Thus, individual team
Interpersonal KSAs
Collaborative Problem Solving KSAs
R4. The KSA to identify situations requiring participative group problem solving and to utilize the proper degree and type of participation.
Participation by all team members in every decision may not always be wise.
The degree of participation should vary as a function of the characteristics of the decision, importance of team acceptance, simplicity of the decision, availability of information, and capability of the members

10. Interpersonal KSAs Collaborative Problem Solving KSAs
1010
Another example is the nominal group technique,
where ideas are generated privately and team meetings only involve presenting
and clarifying ideas. Decisions are made by private voting rather than trying
to publicly resolve differences (Delbecq, Van de Ven & Gustafson, 1975;
Rogelberg, Barnes-Farrell & Lowe,
Interpersonal KSAs
Collaborative Problem Solving KSAs
R5. The KSA to recognize the obstacles to collaborative group problem solving and to implement appropriate corrective actions.
problems of groupthink and conformity.
Many techniques have been developed for avoiding obstacles to group problem solving.
They usually involve restricting interactions so as to limit negative team influences while maximizing positive ones.

11. 1111
Staffing Work Teams: Development and
Validation of a Selection Test for
Teamwork Settings
KSA Test
were used to develop a paper-and-pencil test of teamwork situations.
Standard test construction techniques
35 multiple-choice items on hypothetical teamwork situations.
Two validation studies
N=70 e N=72

13. Situational questions
1313
Respondents indicated what they would do in each situation by selecting
from four multiple-choice alternatives. The alternatives were written to have similar
social desirability to reduce faking. The correct alternative was determined
from findings in the research literature. Questions were dichotomously scored (0/1)
and total scores were the sums of correct answers
Situational questions
Were used for three reasons
high validity in other selection instruments, especially structured interviews
when written in terms of job situations, such questions can have high face validity.
And third, there is some evidence from the research on tacit intelligence that such questions may allow for the measurement of attributes independent of general mental ability

14. Teamwork KSA Test Convergent Validity
1414
The Teamwork Test may have a large mental ability
component because it is a knowledge test and uses a paper-and-pencil format. As
such, a high degree of convergence would be expected with traditional employment
aptitude tests based on the positive correlations observed among
Teamwork KSA Test
Convergent Validity
H1: The Teamwork Test should correlate positively with traditional employment aptitude tests.
May have a large mental ability component because it is a knowledge test and uses a paper-and-pencil format.

15. Teamwork KSA Test Criterion-Related Validity
1515
Teamwork KSA Test
Criterion-Related Validity
H2: The Teamwork Test should correlate positively with measures of individual team member performance.
Consistent with the individual level focus

16. Teamwork KSA Test Criterion-Related Validity
Further, because the emphasis is on teamwork rather than taskwork performance,
the Teamwork Test should be more highly correlated with teamwork
performance than should the employment aptitude tests, whereas the aptitude tests
should be more highly correlated with taskwork performance than should the
Teamwork Test. Thus, the following hypotheses not only address criterion-related
validity, but also construct validity in that they help explain the underlying relationships
between the predictors and the various constructs of performance.
Teamwork KSA Test
Criterion-Related Validity
H3a: The correlation between the teamwork performance ratings and the Teamwork Test should be greater than the correlation between the teamwork performance ratings and the employment aptitude tests.
- the aptitude tests should be more highly correlated with taskwork performance
1616

17. Teamwork KSA Test Criterion-Related Validity
1717
Further, because the emphasis is on teamwork rather than taskwork performance,
the Teamwork Test should be more highly correlated with teamwork
performance than should the employment aptitude tests, whereas the aptitude tests
should be more highly correlated with taskwork performance than should the
Teamwork Test. Thus, the following hypotheses not only address criterion-related
validity, but also construct validity in that they help explain the underlying relationships
between the predictors and the various constructs of performance.
Teamwork KSA Test
Criterion-Related Validity
H3b: The correlation between the taskwork performance ratings and the employment aptitude tests should be greater than the correlation between the taskwork performance ratings and the Teamwork Test.

18. Teamwork KSA Test Incremental Validity
1818
Because traditional employment aptitude tests are
widely used, readily available, and among the best overall predictors of performance,
they would, thus, constitute an appropriate standard of comparison for a
new selection instrument. Thus,
Teamwork KSA Test
Incremental Validity
H4: The Teamwork Test should have incremental criterion-related validity beyond the employment aptitude tests.

19. Study One pulp mill company
All subjects (n = 70) were current employees applying for the new jobs.
Employment Aptitude Tests: 9
Three measured verbal ability
Three measured quantitative ability
Two measured perceptual ability
The final test measured mechanical ability

20. Study One
An overall aptitude test composite score was calculated by converting the nine individual test scores to z-scores and then averaging.
Five items were created for the study:
three reflected teamwork performance (self- management, team contribution, and communication)
two reflected taskwork performance (technical knowledge and learning orientation).
An overall performance measure was also obtained by combining all items

21. Study One
Five supervisors provided ratings on all employees, with each supervisor rating 30 employees on average.
To enhance reliability, ratings were provided by multiple independent supervisors.
Items were defined by a brief explanation and coupled with a 5-point scale
ranging from 5--"well above average" (top 20% of employees) to 1--"well below average" (bottom 20% of employees).

22. Study One
Results
H1: The Teamwork Test should correlate positively with traditional employment aptitude tests.
H1 Supported
H2: The Teamwork Test should correlate positively with measures of individual team member performance.
Teamwork Test correlates with ratings of teamwork performance (r = .44), taskwork performance (r = .56), and overall performance (r = .52).

23. Study One
Results
H3a: The correlation between the teamwork performance ratings and the Teamwork Test should be greater than the correlation between the teamwork performance ratings and the employment aptitude tests.
As predicted, the correlation between teamwork performance and the Teamwork Test is larger than the correlation between teamwork performance and the employment aptitude tests (r = .44 versus .33), but this difference is only significant atp < .10 (t = 1.62, df= 67, one-tailed), thus providing marginal support for Hypothesis 3a.

24. Study One
Results
H3b: The correlation between the taskwork performance ratings and the employment aptitude tests should be greater than the correlation between the taskwork performance ratings and the Teamwork Test.
the correlation between taskwork performance and the aptitude test composite is not significantly larger than the correlation between taskwork performance and the Teamwork Test (r = .60 versus .54; t = 0.67, df= 67, ns), thus not supporting Hypothesis 3b.

25. Study One
Results
H4: The Teamwork Test should have incremental criterion-related validity beyond the employment aptitude tests.
There is a significant increase in explained variance by the Teamwork Test beyond the aptitude test composite for both teamwork laerformance (incremental R = .08) and overall job performance (incremental R2= .06), but not for taskwork performance (incremental R 2 = .01). Thus, Hypothesis 4 is largely supported.

26. Study Two two northeastern cardboard box plants
Unlike Study 1, participation was voluntary and not linked to any work-related outcomes
n = 72 subjects
Employment Aptitude Tests
only two of the nine employment aptitude tests from Study 1 were used: vocabulary and math problem solving.
they had the highest criterion-related validities in Study 1, they represented the most common constructs in traditional selection test batteries (i.e., verbal and math), and they were efficient in terms of time usage.

27. Study Two
The measure was a slightly expanded version of the ratings used in Study 1.
The supervisor teamwork rating included five items (resolving conflicts, collaborative behaviors, interpersonal communication, goal setting and performance management, and coordinating and planning).
The supervisor taskwork rating included three items (technical knowledge depth, technical knowledge breadth, and learning orientation).
A supervisor overall rating was also obtained by combining all items.

28. Study Two Peer nominations were used rather than peer ratings;
It asked the nominator to identify the top one-third preferred peers in each of three categories:
(1) "For promoting good working relationships;" (2) "For helping take charge and staying focused;" and (3) "For technical expertise and know-how."
These categories were chosen to reflect the interpersonal, self- management, and task performance dimensions, respectively
A peer overall nomination was obtained by averaging all three categories.
Peer nominations were used rather than peer ratings
because of their presumed higher validity and reliability (Kane & Lawler, 1978),
and because the unionized setting discouraged use of a measure that might ask
employees to make negative ratings of other employees.
The nomination form
contained names of co-workers from the same shift who were also participating in
the study. It asked the nominator to identify the top one-third preferred peers in
each of three categories: (1) "For promoting good working relationships;" (2)
"For helping take charge and staying focused;" and (3) "For technical expertise
and know-how." A "Do not know this person well enough" option was also
allowed. These categories were chosen to reflect the interpersonal, self-management,
and task performance dimensions, respectively. Peers were instructed that
they could nominate the same person for more than one category, that everyone
did not have to be nominated, and that their responses were for research purposes
only and would be held strictly confidential (e.g., co-workers, supervisors, or
members of management would not see the responses).

29. Study Two Results H1 supported H2 partially supported
Teamwork Test does not show criterion-related validity for peer nominations of taskwork performance nor for any of the self ratings
Table 4 also shows that Hypothesis 2 is partially supported by positive correlations
between the Teamwork Test and supervisory ratings of teamwork performance
(r = .21), taskwork performance (r = .25), and overall performance (r =
.23), as well as for peer nominations of teamwork (r = .23) and overall performance
(r = .21). However,

30. Study Two Results H3a not supported H3b partially supported
the correlations between teamwork performance and the Teamwork Test are not significantly greater than the correlations between teamwork performance and the employment aptitude tests.
H3b partially supported
The correlation between taskwork performance and the aptitude test composite is significantly greater than the correlation between taskwork performance and the Teamwork Test for both the supervisory ratings and for the self-ratings
Table 4 also shows that Hypothesis 2 is partially supported by positive correlations
between the Teamwork Test and supervisory ratings of teamwork performance
(r = .21), taskwork performance (r = .25), and overall performance (r =
.23), as well as for peer nominations of teamwork (r = .23) and overall performance
(r = .21). However,

31. Study Two Results H4 not supported
Teamwork Test did not show significant incremental validity beyond the aptitude test composite for any of the criterion measures (for the teamwork criterion, supervisory ratingand for the peer nomination).
Conversely, the aptitude test composite did show incremental validity beyond the Teamwork Test for the taskwork criterion measure for both the supervisory Ratings and self ratings
Table 4 also shows that Hypothesis 2 is partially supported by positive correlations
between the Teamwork Test and supervisory ratings of teamwork performance
(r = .21), taskwork performance (r = .25), and overall performance (r =
.23), as well as for peer nominations of teamwork (r = .23) and overall performance
(r = .21). However,

32. Anita C. McClough and Steven G. Rogelberg
Selection in Teams: An Exploration of the Teamwork Knowledge, Skills, and Ability Test
Anita C. McClough and Steven G. Rogelberg

33. Teamwork KSA Test
The teamwork KSA test was designed to predict individual behavior within teams.
we examined this test's relationship with both the behavior of the assigned leader in a team and the behavior of the other team members, respectively.
57 ad hoc student teams (N=227)

34. Teamwork KSA Test
The teamwork KSA test was designed to predict individual behavior within teams.
we examined this test's relationship with both the behavior of the assigned leader in a team and the behavior of the other team members, respectively.
57 ad hoc student teams (N=227)

35. 35
In the original study individual behavior within teams was assessed via global supervisor and peer ratings of teamwork in general.
an assessment of each team member‘s behavior during a specific problem-solving meeting can be conducted by a external evaluator.
Hyphotesis
Hypothesis 1: The teamwork KSA test will correlate positively with a measure of observable teamwork behaviors as rated by independent raters such that high scores on the teamwork KSA test are related to greater individual effectiveness within a team.
In the original study individual behavior within teams was assessed via global supervisor and peer ratings of teamwork in general.

36. 36
In the original study individual behavior within teams was assessed via global supervisor and peer ratings of teamwork in general.
an assessment of each team member‘s behavior during a specific problem-solving meeting can be conducted by a external evaluator.
Hyphotesis
Hypothesis 1: The teamwork KSA test will correlate positively with a measure of observable teamwork behaviors as rated by independent raters such that high scores on the teamwork KSA test are related to greater individual effectiveness within a team.
In the original study individual behavior within teams was assessed via global supervisor and peer ratings of teamwork in general.

37. Teamwork Self-Efficacy37
Self-efficacy for teamwork is a
specific type of self-efficacy and can be defined as how
confident an individual is in performing the behaviors
required in a team setting. Because of the aforementioned
relationship between self-efficacy, effort, and per-
formance, it is likely that an individual's self-efficacy
for teamwork will affect that individual's actual
teamwork behaviors. When an individual is confident
in his or her ability to work in teams (high self-efficacy
for teamwork), then he or she may be more willing to
assert teamwork knowledge, skills and abilities.
Teamwork Self-Efficacy
Potential moderators of the relationship between teamwork knowledge and teamwork behavior.
Self-efficacy is na individual's belief that he or she will successfully perform the behaviors required for a specific task
When an individual is confident in his or her ability to work in teams (high self-efficacy for teamwork), then he or she may be more willing to assert teamwork knowledge, skills and abilities.

38. Teamwork Self-Efficacy38
Self-efficacy for teamwork is a
specific type of self-efficacy and can be defined as how
confident an individual is in performing the behaviors
required in a team setting. Because of the aforementioned
relationship between self-efficacy, effort, and per-
formance, it is likely that an individual's self-efficacy
for teamwork will affect that individual's actual
teamwork behaviors. When an individual is confident
in his or her ability to work in teams (high self-efficacy
for teamwork), then he or she may be more willing to
assert teamwork knowledge, skills and abilities.
Teamwork Self-Efficacy
Potential moderators of the relationship between teamwork knowledge and teamwork behavior.
Self-efficacy is na individual's belief that he or she will successfully perform the behaviors required for a specific task
When an individual is confident in his or her ability to work in teams (high self-efficacy for teamwork), then he or she may be more willing to assert teamwork knowledge, skills and abilities.

39. Measures Teamwork Self-Efficacy39
Measures
Teamwork Self-Efficacy
A revised version of the Personal Efficacy Beliefs Scale (PEBS)
Individual teamwork performance- independent evaluation.
Individual Performance in Teams Scale (IPIT)
Individual teamwork performance-peer ratings.
five dimensions (e.g., participation in the group, interpersonal skills).

40. 40
Procedure
Participants individually completed the teamwork KSA instrument, the self-efficacy for teamwork scale, and a demographic questionnaire during a regular class meeting time.
leaders were chosen randomly
A video camera recorded the participants working
Raters watched videotapes of each team and then rated each team member.

41. Results Team members (n=170) H1 and H2 were supported41
Teamwork KSA and individual performance in teams.
To test Hypotheses 1 and 2, a Pearson product moment
correlation was performed between teamwork KSA
scores and IPIT scores, and teamwork KSA scores and
peer ratings. As expected, we found that the teamwork
KSA test successfully predicted individual member
behavior within teams as indexed by external raters (r
. .31, p ) and peers (r . .34, p ) such that
higher scores on the teamwork KSA test related to
greater individual effectiveness within the team.
Exploratory analyses with polynomial regression were
conducted to determine whether a curvilinear
relationship (e.g., U-shaped) existed between the
teamwork KSA scores and the two criteria measures,
Results
Team members (n=170)
H1 and H2 were supported
H3 was not supported
suggesting that self-efficacy for teamwork does not moderate the relationship between teamwork KSA scores and IPIT.

42. Results Team members (n=170) H1 and H2 were supported42
Teamwork KSA and individual performance in teams.
To test Hypotheses 1 and 2, a Pearson product moment
correlation was performed between teamwork KSA
scores and IPIT scores, and teamwork KSA scores and
peer ratings. As expected, we found that the teamwork
KSA test successfully predicted individual member
behavior within teams as indexed by external raters (r
. .31, p ) and peers (r . .34, p ) such that
higher scores on the teamwork KSA test related to
greater individual effectiveness within the team.
Exploratory analyses with polynomial regression were
conducted to determine whether a curvilinear
relationship (e.g., U-shaped) existed between the
teamwork KSA scores and the two criteria measures,
Results
Team members (n=170)
H1 and H2 were supported
H3 was not supported
suggesting that self-efficacy for teamwork does not moderate the relationship between teamwork KSA scores and IPIT.

43. Conclusions
First, the teamwork KSA test was able to predict individual team member behavior
Second, it appears as if the teamwork KSA test's effectiveness in predicting team member behavior is not dependent upon members’ teamwork self-efficacy.
Third, the validity of the teamwork KSA test was found to generalize beyond the type of team it was designed for, self-directed work teams (i.e., ad hoc temporary student teams).
Finally, in addition to peer ratings, the validity of the teamwork KSA test was established using a new direct and `objective' index of member behavior (the IPIT).