Dissertation Subject Guide

Quantitative data analysis involves processing and interpreting numerical data to draw meaningful conclusions. Here are some key steps in quantitative data analysis:

1. Data Preparation:

   • Clean and organize the data.
   • Handle missing values and outliers.
   • Transform variables if needed (e.g., normalization, standardization).

2. Descriptive Statistics:

   • Calculate measures like mean, median, mode, and standard deviation.
   • Create histograms, box plots, or scatter plots to visualize data distributions.

3. Inferential Statistics:

   • Perform hypothesis tests (e.g., t-tests, ANOVA) to compare groups.
   • Calculate confidence intervals.
   • Assess relationships (correlation, regression).

4. Statistical Software:

   • Use tools like R, Python (with libraries like pandas, numpy, and scipy), or SPSS for analysis.

META-ANALYSIS

 When conducting data analysis for a meta-analysis, follow these steps:

1. Data Extraction:

   • Collect relevant data from each study, including effect sizes, sample sizes, and other relevant statistics.
   • Ensure consistency in data extraction across studies.

2. Effect Size Calculation:

   • Compute effect sizes (e.g., Cohen’s d, odds ratio, correlation coefficient) for each study.
   • Standardize effect sizes to facilitate comparison.

3. Forest Plot:

   • Create a forest plot to visualize effect sizes and confidence intervals for each study.
   • Identify the overall effect size (pooled estimate) and its precision.

4. Heterogeneity Assessment:

   • Evaluate heterogeneity among studies using statistical tests (e.g., Cochran’s Q, I²).
   • High heterogeneity may require subgroup analyses or sensitivity analyses.

5. Fixed-Effect or Random-Effects Model:

   • Choose an appropriate model (fixed-effect or random-effects) based on heterogeneity.
   • Fixed-effect assumes a common effect size, while random-effects accounts for variability.

6. Publication Bias:

   • Assess publication bias using funnel plots or statistical tests (e.g., Egger’s test).
   • Adjust for bias if necessary.

Remember that meta-analysis requires careful consideration of study quality, study design, and statistical assumptions.

https://training.cochrane.org/handbook/current/chapter-10

NARRATIVE SYNTHESIS

Data analysis for narrative synthesis involves collating and organizing study findings from different studies in a review. Unlike meta-analysis, which uses statistical methods, narrative synthesis relies on textual descriptions to integrate results. Here are some key steps:

1. Collate Findings: Describe the main features of each study, including context, validity, and differences in characteristics.

2. Structured Tabulation: Use tables and graphs to display results and highlight variations across studies.

3. Transparency: Promote transparency by justifying decisions and pre-specifying the synthesis approach in the review protocol

https://cccrg.cochrane.org/sites/cccrg.cochrane.org/files/uploads/AnalysisRestyled.pdf

Braun and Clarke (2006) thematic analysis method is a process consisting of six steps:

1. becoming familiar with the data
2. generating codes
3. generating themes
4. reviewing themes
5. defining and naming themes
6. locating exemplars

Braun, V. and Clarke, V. (2006) ‘Using thematic analysis in psychology’, Qualitative research in psychology, 3(2), pp. 77–101. Available at: https://doi.org/10.1191/1478088706qp063oa.
 

Watch these youtube lecture videos on Thematic Analysis, presented by Victoria Clarke in 2021:

Thematic Analysis Part 1: What is Thematic Analysis?

Thematic Analysis Part 2: Thematic Analysis is uniquely flexible

Thematic Analysis Part 3: Six phases of reflective thematic analysis

Thematic Analysis Part 4: Avoiding common problems