Statistical Analysis: Customer Satisfaction Study
Research Question: Does customer support response time predict customer satisfaction scores?
Sample Size: N = 450 survey responses
Software: SPSS v29
Methods:
A linear regression analysis was conducted to examine the relationship between response time (independent variable, measured in hours) and satisfaction score (dependent variable, 1-10 scale). Assumptions tested included linearity, independence of errors (Durbin-Watson = 1.94), homoscedasticity, and normality of residuals.
Results:
Response time significantly predicted satisfaction scores, F(1, 448) = 124.67, p < .001, R² = .218. Response time accounted for 21.8% of the variance in satisfaction. The regression coefficient was negative (B = -0.42, SE = 0.04, β = -0.47, p < .001), indicating that each additional hour of response time decreased satisfaction by 0.42 points.
Table 1. Regression Results
| Variable | B | SE | β | p-value |
|---|---|---|---|---|
| (Intercept) | 8.64 | 0.18 | - | <.001 |
| Response Time (hrs) | -0.42 | 0.04 | -0.47 | <.001 |
Statistical Tests We Perform:
- Descriptive statistics (mean, median, SD, range)
- T-tests (independent, paired, one-sample)
- ANOVA (one-way, two-way, repeated measures)
- Correlation (Pearson, Spearman)
- Regression (linear, multiple, logistic)
- Chi-square tests for categorical data