Research Results
scepe41
SharonReedresultspptslides.pptxSoon-to-be Dr. Sharon Reed
Results ppt Slides for your Consideration
Research Questions
Research Question 1 (RQ1)
Is there a difference in aims or goals emphasis, as a leadership behavior, between non-profit human services managers with and without human services degrees?
Research Question 2 (RQ2)
Is there a difference in personal backing, as a leadership behavior, between non-profit human services managers with and without human services degrees?
Research Question 3 (RQ3)
Is there a difference in work easing or facilitation, as a leadership behavior, between non-profit human services managers with and without human services degrees?
This study was guided by three research questions. They correspond to the three leadership behaviors of interest (aims or goals emphasis; personal backing; and work easing or facilitation). Associated hypotheses are listed in sections that present the results for the research questions.
2
Data Handling
Percentages rounded to whole numbers
Data were analyzed with SPSS v 25
Statistical significance alpha = .050
Reliability statistics w Cronbach’s alpha
Aims emphasis SS
Personal Backing SS
Work Easing SS
Percentages were rounded off to whole numbers and may not add up to precisely 100%. Data were analyzed with SPSS v 25, which is dedicated statistical software. Statistical significance was set at alpha = .050.
Reliability
Reliability statistics (i.e., internal consistency) were run on conceptually-similar survey items with Cronbach’s alpha for the three leadership behaviors of interest (aims or goals emphasis; personal backing; and work easing or facilitation). Cronbach’s alpha statistics range in value from 0 to 1. The closer Cronbach’s alpha is to 1, the greater the reliability of the database. Indices of .70 or higher reflect an adequately reliable database (Gliner & Morgan, 2000).
Summated Scale Scores
A summated scale was generated for each of the three leadership behaviors of interest (aims or goals emphasis; personal backing; and work easing or facilitation). A summated scale is a single empirical measure that represents multiple aspects of a construct in one variable (Hair et al., 2010). Deriving a single measure from several related aspects decreases the measurement error in the original data points, which increases data reliability and validity as well as parsimony in the overall number of variables (Hair et al., 2010). Because each summated score was a mean, its possible values ranged 1 – 7 (1 = strongly disagree, 7 = strongly agree) like the Likert scale used to measure responses to survey items.
Aims emphasis. Two survey items measured the leadership dimension of “aims emphasis.” An example of a survey item that measured aims emphasis is: “My manager makes sure subordinates have clear goals to achieve.” Internal consistency was high, Cronbach’s alpha = .74. Because of high internal consistency, a summated scale (SS) was generated, using the mean response of the related items. It is hereafter called the Aims Emphasis SS.
Personal backing. Three survey items measured the leadership dimension of “personal backing.” An example of a survey item that measured personal backing is: “My manager is concerned about his/her subordinates as people.” Cronbach’s alpha showed that internal consistency was high, alpha = .84. Because of high internal consistency, a summated scale (SS) was generated, using the mean response of the related items. It is hereafter called the Personal Backing SS.
Work easing. Four items measured the leadership dimension of “work easing.” An example of a survey item that measured work easing is: “My manager helps his/her subordinates solve work-related problems.” Cronbach’s alpha showed that internal consistency was high, alpha = .94. Because of high internal consistency, a summated scale (SS) was generated, using the mean response of the related items. It is hereafter called the Work Easing SS.
3
Demographic Characteristics
113 individuals completed surveys
10 participants (cases 4, 14, 20, 21, 33, 45, 62, 78, 101, 105) failed to provide information on their managerial educational background, leadership style, or both and were eliminated from further analyses, N = 103 participants
Modal participant was a married Caucasian woman in her 40s who had worked for her current organization for ten years and for her current manager for five years
Demographic Characteristics
This section presents demographic variables and characterizes the modal participant. A total of 113 individuals completed surveys. Of those, 10 participants (cases 4, 14, 20, 21, 33, 45, 62, 78, 101, 105) failed to provide information on their managerial educational background, leadership style, or both. They were eliminated from further analyses, N = 103 participants. The demographic results in this section show that the modal participant was a married Caucasian woman in her 40s who had worked for her current organization for ten years and for her current manager for five years.
There were three times as many participants whose managers did not hold human service degrees, n = 75 participants, 73%, as participants who managers held human services degrees, n = 28 participants, 28%. The 3-to-1 ratio of non-degree-holders to degree-holders is illustrated on Figure 1.
4
Human Services Degree Holders
NO
YES
Figure 1. Numbers of participants by human services degrees.
There were three times as many participants whose managers did not hold human service degrees, n = 75 participants, 73%, as participants who managers held human services degrees, n = 28 participants, 28%. The 3-to-1 ratio of non-degree-holders to degree-holders is illustrated on Figure 1.
5
Gender and Ethnicity
Figure 2. Numbers of participants by ethnicity.
There was a gender ratio of 4-to-1 women to men; women 80%, n = 82 female participants, men 20%, n = 21 male participants. Four of the 103 participants did not provide ethnic information. Of the 99 participants who provided ethnic information, illustrated on Figure 2, the majority of the participants were Caucasian, 77%, n = 76 participants. The next largest group was African American, 12%, n = 12 participants. The remaining ethnicities were each represented by four or fewer participants.
6
Marital Status
Figure 3. Numbers of participants by marital status.
Figure 3 illustrates the participants by marital status. Half were married, 50%, n = 51 participants. Approximately half were single, 41%, n = 42. Ten participants labeled their marital status as something other than married or single, 10%.
7
Age and Time with Organization & Manager
Average age: early 40’s
M = 43.38 years old, SD = 13.13, min = 21 years, max = 70 years
Average 10 years working for current organization
M = 9.84 years, SD = 8.02, min = 1 year, max = 31 years
Average 5 years working for current manager
M = 4.87 years, SD = 5.48, min = 1 year, max = 30 years
Participants were in their early 40’s on average, M = 43.38 years old, SD = 13.13, but represented a broad range of ages, min = 21 years, max = 70 years. They had been working for their current organization nearly ten years on average, M = 9.84 years, SD = 8.02, though again represented a broad range of time on the current job, min = 1 year, max = 31 years. They had worked for their current manager about half as long as they had been employed by the organization, M = 4.87 years, SD = 5.48. They again, however, reported a broad range, min = 1 year, max = 30 years.
8
Human Services Education vs Experience
Chi-square H0: The association between human services degrees and professional background in human services was not statistically significant.
Association significant, X2 (1, 102) = 9.53, p = .002; null rejected
Figure 4. Cross-tabulation between human services degree held (yes or no) and manager’s professional background in human services (yes or no).
Each participant was a member of one of two groups: those whose managers held a human services degree and those whose managers did not hold a human services degree. Participants were also asked on the survey whether or not their manager’s professional background included human services experience, regardless of formal education.
Figure 4 shows the cross-tabulation of human services degrees (formal education) and professional background (direct experience) among managers. In total, 28 managers held human services degrees. All 28 of them had a professional background in human services as well. The other 74 managers did not hold a human services degree. In a 2-to-1 ratio, twice as many non-degree-holding managers had professional backgrounds in human services as did not have professional backgrounds in human services.
A chi-square test was run to determine whether the association between holding a human services degree and having a professional human services background was statistically significant. The hypotheses were:
H0: The association between human services degrees and professional background in human services was not statistically significant.
H1: The association between human services degrees and professional background in human services was statistically significant.
Results of the chi-square indicated that the association between human services degrees and professional background in human services was statistically significant, X2 (1, 102) = 9.53, p = .002, Φ = .33, p = .001. The null hypothesis was rejected. Adjusted residuals, listed on Table 1, showed that there were significantly fewer degree holders without human services backgrounds than expected but also significantly more non-degree-holders without human services backgrounds than expected.
9
Descriptive Statistics for Summated Scales and Managerial Effectiveness
“somewhat agree” to “agree”
“moderate to maximal”
Table 2
Descriptive Statistics for Summated Scales and Managerial Effectiveness
Descriptive Statistics for Summated Scales and Managerial Effectiveness
Recall that the Likert scale of agreement used to measure the three leadership behaviors of interest (aims or goals emphasis; personal backing; and work easing or facilitation) had a 7-point scale (1 = strongly disagree, 7 = strongly agree). Also recall that summated scale scores were the means of each participant’s responses to conceptually-related items. Table 2 shows the descriptive statistics for the three summated scales and a measure of manager effectiveness.
Means for all three summated scales fell between 5 and 6 in value. Scores between 5 and 6 reflect responses between “somewhat agree” to “agree.” The minimum score for all three summated scales was one, indicating that at least participant strongly disagreed with all of the survey items used to measure a particular summated scale. Alternatively, the maximum score for all three summated scales was 7, indicating that at least one participant strongly agreed with all of the survey items used to measure a particular summated scale.
The fourth measure whose descriptive statistics are listed on Table 2 summarizes overall managerial effectiveness. These data emerged from responses to the survey question, “How effective is your manager? Pleases choose one number between 1 and 10, with 1 = not at all effective, 10 = maximally effective.” On average, participants rated managerial effectiveness between 7 and 8, indicating that average perceptions were that managerial effectiveness fell between moderate and maximal.
10
Correlations
| V1 | V2 | V3 | V4 | V5 | V6 | V7 | |
| V1 Personal Backing SS | .82** | .84** | .71** | .13 | -.06 | -.06 | |
| V2 Aims Emphasis SS | .71** | .86** | .75** | .14 | -.02 | -.07 | |
| V3 Work Easing SS | .84** | .92** | .81** | .19 | -.05 | -.02 | |
| V4 Managerial Effectiveness | .71** | .69** | .77** | .08 | .02 | .01 | |
| V5 Years w Current Manager | .21 | .19 | .16 | .15 | .59** | .45** | |
| V6 Year w Current Organization | -.20 | -.18 | -.23 | -.20 | .66** | .48** | |
| V7 Age | -.03 | -.09 | -.04 | .04 | .35 | .43* |
Correlations
This section shows the results of inspecting correlations to look for covariates and sufficient correlations among dependent variables to establish the analytical approach to answering RQs 1-3. The analytical plan was based on correlations: analyze RQs 1-3 with separate independent t tests in the absence of covariates; analyze RQs 1-3 with 1x2 ANCOVA tests in the presence of covariates; analyze RQs 1-3 with a MANOVA test if the dependent variables (leadership behaviors: aims or goals emphasis; personal backing; and work easing or facilitation) were sufficiently correlated; or analyze RQs 1-3 with a MANCOVA test the dependent variables were sufficiently correlated and there were covariates.
The patterns of the correlations among participants whose managers did and did not hold a human services degree are shown on Table 3. Correlations above the diagonal reflect managers without human services degrees, n = 73-75 participants. Correlations below the diagonal reflect managers with human services degrees, n = 27-28 participants.
The patterns across the two groups were identical. Four patterns emerged. One, all three summated scales (V1, V2, and V3) were strongly and positively correlated to each other. This justified a MANOVA test. Two, all three summated scales (V1, V2, and V3) were also strongly and positively correlated with the numeric estimate of the manager’s overall effectiveness (V4). This justified a MANCOVA test. Three, there were strong and positive correlations between the years spent working for the current manager, years spent working for the current human services organization, and the participant’s age (V5, V6, and V7) . Four, these three demographic variables (V5, V6, and V7) did not correlate significantly with the summated scales (V1, V2, and V3) or managerial effectiveness (V4). These demographic variables were therefore not used as covariates.
11
Results for Research Questions 1-3
MANCOVA test addressed 3 RQs at once in a single analysis
Covariate H0: Managerial effectiveness was not a significant covariate.
Managerial effectiveness significant covariate, Wilks Lambda = .35, F(3, 96) = 60.29, p < .001; null rejected
Multivariate H0: The difference in the linear combination of the three leadership behaviors of interest between managers with and without human services degrees was not statistically significant.
Group difference in leadership not statistically significant, Wilks Lambda = .99, F(3, 96) = 0.37, p = .775, null retained
Results for Research Questions 1-3
Correlations listed on Table 3 justified a MANCOVA test that addressed all three research questions at once in a single analysis. This section presents those results. RQ1 was, Is there a difference in aims or goals emphasis, as a leadership behavior, between non-profit human services managers with and without human services degrees? RQ2 was, Is there a difference in personal backing, as a leadership behavior, between non-profit human services managers with and without human services degrees? RQ3 was, Is there a difference in work easing or facilitation, as a leadership behavior, between non-profit human services managers with and without human services degrees?
In the MANCOVA, the covariate was the managerial effectiveness. The independent variable was human services degree, with two levels (yes, manager held a human services degree; no, manager did not hold a human services degree). The multivariate dependent variable was the linear combination of the three leadership behaviors of interest (aims or goals emphasis; personal backing; and work easing or facilitation). MANCOVA first tests a set of multivariate hypotheses. The multivariate hypotheses were:
Covariate H0: Managerial effectiveness was not a significant covariate.
Covariate H1: Managerial effectiveness was a significant covariate.
Multivariate H0: The difference in the linear combination of the three leadership behaviors of interest between managers with and without human services degrees was not statistically significant.
Multivariate H1: The difference in the linear combination of the three leadership behaviors of interest between managers with and without human services degrees was statistically significant.
The data met the assumption of equality of covariance matrices, Box’s M = 16.44, F(6, 16609) = 2.62, p = .021. The data also met the assumption of equality of error variances, Aims Emphasis SS Levene’s F(1, 99) = 0.52, p = .474; Personal Backing SS Levene’s F(1, 99) = 0.12, p = .735; Work Easing SS Levene’s F(1, 99) = 0.95, p = .331.
Results of the multivariate portion of the MANCOVA showed that the managerial effectiveness was a significant covariate, Wilks Lambda = .35, F(3, 96) = 60.29, p < .001; the covariate null hypothesis was rejected. The impact of the overall managerial effectiveness was very strong, partial eta2 = .65.
When the impact of the overall managerial effectiveness was factored out, the difference in the linear combination of the three leadership behaviors of interest between managers with and without human services degrees was not statistically significant, Wilks Lambda = .99, F(3, 96) = 0.37, p = .775. The multivariate null hypothesis was retained.
12
Group Difference in Aims Emphasis
Impact of human services degree very strong, partial eta2 = .45
NO
YES
Figure 5. Mean Aims Emphasis SS across managers with and without a human services degree.
Figure 5 illustrates mean ratings for Aims Emphasis across the two groups. The average rating for the Aims Emphasis SS was slightly higher among managers who held human services degrees, M = 5.91, SD = 0.98, than it was among managers who did not hold human services degrees, M = 5.45, SD = 1.27. Although the difference in means was non-significant, as per the MANCOVA results, the impact of human services degrees on this dimension of leadership was very strong, partial eta2 = .45.
13
Group Difference in Personal Backing
Impact of human services degree very strong, partial eta2 = .50
NO
YES
Figure 6. Mean Personal Backing SS across managers with and without a human services degree.
Figure 6 illustrates mean ratings for Personal Backing across the two groups. The average rating for the Personal Backing SS was slightly higher among managers who held human service degrees, M = 5.71, SD = 1.40, than it was among managers who did not hold human services degrees, M = 5.15, SD = 1.45. Although the difference in means was non-significant, as per the MANCOVA results, the impact of human services degrees on this dimension of leadership was very strong, partial eta2 = .50.
14
Group Difference in Work Easing
Impact of human services degree very strong, partial eta2 = .42
NO
YES
Figure 7. Mean Work Easing SS across managers with and without a human services degree.
Figure 7 illustrates mean ratings for Work Easing across the two groups. The average rating for the Work Easing SS was slightly higher among managers who held human services degrees, M = 5.79, SD = 1.28, than it was among managers who did not hold human services degrees, M = 5.30, SD = 1.49. Although the difference in means was non-significant, as per the MANCOVA results, the impact of human services degrees on this dimension of leadership was very strong, partial eta2 = .42.
15
Ideal Non-profit Leadership
Authoritarian: Leaders who have an authoritarian management style have a high concern for task and little concern for people.
Compromise management: Leaders who have a compromise management style have a moderate concern for tasks and a moderate concern for people.
Transformational leadership: Leaders who have a transformational management style are involved with feelings, morals, values, and long–term goals.
Situational leadership: Leaders who have a situational leadership management style are rooted in the communication of relationship activities and task activities, in addition to employees’ willingness or maturity in favor of performing a particular task.
Finally, participants were asked to identify the ideal non-profit leadership by choosing from an array of four leadership styles on two survey questions, defined as follows. Authoritarian: Leaders who have an authoritarian management style have a high concern for task and little concern for people. Compromise management: Leaders who have a compromise management style have a moderate concern for tasks and a moderate concern for people. This kind of manager attempts to stabilize the importance of completing the task while connecting to the subordinates and keeping them pleased since the team is usually effective and they get along with the manager. Compromise managers are capable to eliciting respect from subordinates because they listen to and incorporate subordinates’ ideas. Transformational leadership: Leaders who have a transformational management style are involved with feelings, morals, values, and long–term goals. Situational leadership: Leaders who have a situational leadership management style are rooted in the communication of relationship activities and task activities, in addition to employees’ willingness or maturity in favor of performing a particular task.
Responses to the two survey questions were parsed out by group (human services managers with and without human services degrees) to see if and how they differed.
16
Ideal Leadership Style for a Non-Profit Manager
Non-significant, X2 (2, 102) = 1.22, p = .543, Cramer’s V = .11; null retained
Figure 8. Percentages of participants in three ideal manager leadership styles.
Ideal Leadership Style for a Non-Profit Manager
One of the two survey questions was, “What is the ideal leadership style for a non-profit manager?” Figure 8 illustrates the percentages of participants in the two groups across ideal manager leadership styles. Table 4 lists the percentages and corresponding numbers of participants.
The majority of both types of participants (i.e., whose managers did or did not hold a human services degree) identified compromise management as the ideal leadership style for non-profit managers (see Table 4). The type of style that garnered the next highest percentage of votes in both groups was transformational leadership. Finally, Figure 8 and Table 4 show that situational leadership garnered the fewest votes. No one identified authoritarian leadership as the ideal style for non-profit managers.
Table 4 shows that the percentages of participants who voted for the three leadership styles were close in value but not identical. A chi-square test of independence was therefore run to see if the percentages of votes for the ideal leadership style differed by testing the significance of the association between the cross-tabulated variables. The hypotheses were:
H0: The association between participant type (manager did or did not hold a human services degree) and preferences for the ideal non-profit manager leadership style was not statistically significant.
H1: The association between participant type (manager did or did not hold a human services degree) and preferences for the ideal non-profit manager leadership style was statistically significant.
Results of the chi-square showed that the association between participant type (manager did or did not hold a human services degree) and preferences for the ideal non-profit manager leadership style was not statistically significant, X2 (2, 102) = 1.22, p = .543, Cramer’s V = .11. The null hypothesis was retained.
17
Best Leadership Style for a Non-profit Organization
Non-significant, X2 (3, 103) = 4.58, p = .205, Cramer’s V = .21; null retained
Figure 9. Percentages of participants in four ideal leadership types for non-profit organizations.
Best Leadership Style for a Non-profit Organization
The other survey question was, “Which leadership style is the best for a non-profit organization? Figure 9 illustrates the percentages of participants in the two groups across four ideal manager leadership styles. Table 4 lists the percentages and corresponding numbers of participants.
Figure 9 shows that one person in the group whose managers did not hold a human services degree choose authoritarian for the ideal leadership type of non-profit organizations. Otherwise, the majority of both types of participants (i.e., whose managers did or did not hold a human services degree) said that compromise management was the ideal leadership type for non-profit organizations (see Table 4). That was followed by transformational leadership. Finally, Figure 9 and Table 4 show that situational leadership garnered the fewer votes and authoritarian leadership only one vote.
Table 4 again shows that the percentages of participants who voted for the four leadership type were somewhat close in value but not identical. A chi-square test of independence was again run to see if the percentages of votes differed. The hypotheses were:
H0: The association between participant type (manager did or did not hold a human services degree) and preferences for the ideal leadership type for non-profit organizations was not statistically significant.
H1: The association between participant type (manager did or did not hold a human services degree) and preferences for the ideal leadership type for non-profit organizations was statistically significant.
Results of the chi-square showed that the association between participant type (manager did or did not hold a human services degree) and preferences for the ideal leadership type for non-profit organizations was not statistically significant, X2 (3, 103) = 4.58, p = .205, Cramer’s V = .21. The null hypothesis was retained.
18
