Public Sector Commission (PSC) Survey Analysis
Role: Quantitative Research Analyst
Team Size: Independent
Tools: R (ggplot2, psych, dplyr, stringr), GitHub , SPSS
Duration: 7 days
My Role
Cleaned survey data from 3,800+ public sector employees across 11 agencies.
Conducted factor analysis in SPSS and R to uncover hidden themes in employee perceptions.
Identified three key factors shaping organizational culture: leadership confidence, job fit & autonomy, and work-life balance.
Ensured validity and reliability of findings through KMO, Bartlett’s Test, Cronbach’s Alpha, and composite reliability.
Overview
This analysis focuses mainly on employees’ personal feelings towards their respective agencies and dives into whether these employees want to stay or leave. (Question ID: A1a – A4avii). This survey represents employees from 11 public sector organizations in Western Australia. This study highlights internal factors such as job satisfaction, leadership and work-life balance.
Problem Statement
Public sector agencies lack a clear understanding of how employees perceive leadership, job fit, and work-life balance. Without these insights, it is difficult to identify the drivers of job satisfaction and retention.
Figures
Satisfaction With Career Progression
40%
Workers planning on staying in their agency
agree
disagree
unsure
neutral
There is a clear contrast in satisfaction with career progression between the two groups.
Nearly 60% of respondents who plan to stay with their agency report being satisfied with their career progression. In contrast, about 60% of those who plan to leave indicate dissatisfaction with their career progression in some form.
This suggests that perceived career advancement opportunities may play a significant role in an employee’s decision to remain with or leave an agency.
When respondents don't have a strong personal attachment to their agency they also don't believe that have a strong work life balance opposite for previous figure. While respondents who have a personal attachment to their agency tend to overwhelming agree that they have a good work life balance.
This suggests that work-life balance has an influence on the thoughts and opinions of respondents and with more work life balance employees feel more attached and fulfilled to their agency.
agree
disagree
unsure
neutral
60%
vs
agree
disagree
don’t know
50%
Employees Feelings of Work Life Balance
Workers Who Have a Strong Connection to Agency
Factor Analysis Motivation
To simplify the data and identify underlying themes factor analysis helps reduce variables into a smaller set of meaningful factors. This helps us understand what hidden feelings, as a collective, employees may feel toward their agency.
Methodology
Factor Loading: Correlation between a variable, in this case survey items, and a given factor. Shows the extent to which the two are related
Rotation Methods: Varimax forces factors apart (assumes independence), while Oblimin lets them overlap (assumes real-world connections). Use component correlation matrix to see which method to use.
Criteria for Factor Selection: scree plot elbow method
Tables and Graphs
Interpretation: Plot levels after 3 components so this tells us that the rest of the components ( > 3 ) don’t explain that much variance.
Component Correlation Matrix
1
2
3
Component
1
2
3
1.000
-.650
0.371
1.000
-.650
0.371
1.000
-.650
0.371
Interpretation: Oblimin rotation was used because the first 2 components were moderately/highly correlated with each other (r = -0.65). This suggests that the constructs are not statistically independent which makes oblique rotation the better choice for this analysis.
Factors
Factor 1: Captures employees’ emotional attachment to their respective agency and their confidence in the agencies leadership. High scores show us that employees have a strong connection to their agency the negative score A4 (It is likely that you will leave your agency within:..) shows us that these employees are not likely to leave their agency. The items with the highest loadings are variables all involving employee pride.
Factor 2: Captures employee’s feelings about transparency/clarity in their job responsibilities, as well as their understanding of how their work contributes to their agencies objectives. These factor scores are all negative suggesting that there was lower agreement to these.
Factor 3: Captures employee’s opinions on their work-life balance within their agency. High scores on this factor tells us that employees feel their agency offers flexible work options and has a healthy culture that focuses on the well-being of their employees.
Factor Analysis Advanced Overview and Summary of Methods
Factor analysis was conducted using SPSS and R. The aim of this analysis was to uncover any latent themes that would influence employees’ responses. Since Likert-scale data is ordinal in nature the sample size for each item was checked to verify sufficient responses. Several measures were used to assess the data’s suitability for factor analysis, including checking the correlation matrix and determinant score to check multicollinearity, the Kaiser-Meyer-Olkin (KMO) statistic that measures the suitability of data for factor analysis, and Bartlett’s Test of Sphericity which tests the presence of correlation between variables.
The factor extraction method used was principal component analysis, and to decide which factors to keep a scree plot was used in conjunction with the elbow method. The factor rotation used was oblimin due to the moderately high correlation score (r = 0.65) between factors 1 and 2 in the component correlation matrix. This indicates that the factors were not orthogonal. To test consistency and validity Cronbach’s alpha, average variance extracted (AVE) and composite reliability (CR) scores were all taken and interpreted for each factor.
Insights
1
The analysis revealed that public sector employees place the greatest value on organizational structure and leadership, clarity in their roles, and maintaining a healthy workplace culture.
Reflection