Blended-Teaching-Survey

This repository contains the analysis of the "Blended Teaching Experience: Student Satisfaction Survey". The survey was designed to gather feedback from students on their experience with blended learning, including both online and face-to-face components.

The data for this survey was collected using Google Forms, ensuring confidentiality and anonymity for all respondents. The survey included sections on:

• Demographics (program year, comfort with technology, gender, major, current role, and preferred communication media).
• Satisfaction Ratings (using a 5-point Likert scale to assess various aspects of blended learning).

Before visualizing the data in Power BI and running statistical tests in Python, several data wrangling steps were performed:

• ✅ Unpivoted Columns -- Converted wide-format data into a more analysis-friendly long format.
• ✅ Split Columns -- Separated multi-response fields for clearer insights.
• ✅ Added Custom Columns -- Created calculated fields as needed.
• ✅ Replaced Values -- Standardized and cleaned responses for consistency.

To statistically evaluate satisfaction levels, Python was used to conduct two hypothesis tests on the survey data:

1. One-sample t-test (Q9 vs Neutral = 3)
   • Result: The average satisfaction rating was 2.51 (SD = 1.38), which is significantly below the neutral value of 3 (t(234) = −5.49, p< 0.0001 for the one-sided test of μ < 3).
   • Interpretation: Students, on average, reported lower-than-neutral satisfaction with the blended teaching experience.
2. One-way ANOVA (Q9 by Comfort Level with Technology)
   • Result: There was a significant difference among groups (F(2, 232) = 7.73, p < 0.001; η² = 0.062).
   • Group Means:
     • High Comfort: 2.41
     • Medium Comfort: 2.25
     • Low Comfort: 3.08
   • Post-hoc tests (Tukey HSD): Confirmed that students with Low comfort reported significantly higher satisfaction compared to both Medium and High comfort groups.
   • Nonparametric Check: A Kruskal--Wallis test supported this finding (H = 13.72, p = 0.001).

Summary of Insights:

• Overall satisfaction was lower than neutral.
• Comfort with technology plays a role: students who reported lower comfort actually gave higher satisfaction ratings than those with medium or high comfort.

In addition to hypothesis testing, Power BI was used to generate dashboards with:

• Demographic breakdowns (e.g., program year, comfort with technology).
• Satisfaction ratings across different aspects of blended learning.
• Insights into communication media preferences.

• Blended_Teaching_Experience_Student_Survey.docx: The original survey questionnaire.
• Blended_Teaching_Experience_Survey_Results.xlsx: The collected survey responses.
• PowerBI_Report.pbix: The Power BI report (if included).
• blended_satisfaction_tests_summary.csv: Statistical test results from Python.
• blended_satisfaction_group_descriptives.csv: Group-level descriptive statistics.
• blended_satisfaction_tukey_posthoc.csv: Post-hoc test results.
• data/: (optional) Directory containing cleaned and transformed datasets.

1. Clone this repository to your local machine.
2. Open the Power BI report (PowerBI_Report.pbix) to explore interactive dashboards.
3. Review the CSV files generated from Python for detailed statistical analysis.
4. Refer to this README for context on methodology and results.

If you have any questions or suggestions, feel free to open an issue or submit a pull request!