Data Cleaning and Statistical Analysis
A workflow from raw tables through cleaning, descriptive statistics, significance testing and figure export, ready for a paper's results.
Steps
Check the raw data
Excel / LibreOfficeCheck field names, units, missing values, duplicate rows, outliers and data entry errors.
Don't edit the only copy of the raw table; make a copy of the worksheet first.
Batch clean and standardise
OpenRefineUnify category names, fix spelling, merge duplicates and batch-convert text and date formats.
OpenRefine's operation history can be exported, making the cleaning process reproducible.
Prepare the analysis table
R / PythonReshape the data into an analysis format with one sample per row and one variable per column.
Use short English variable names with underscores to avoid frequent errors in code.
Descriptive statistics and visualisation
R / Python / JamoviCompute means, standard deviations and standard errors, and draw box plots, scatter plots and grouped trend charts.
Plot before testing — many outliers and grouping problems are easier to spot in a chart.
Statistical tests
Jamovi / JASP / RChoose t-tests, ANOVA, correlation, regression, non-parametric tests or mixed models according to the study design.
The statistical method must match the experimental design; don't pick one arbitrarily from the software menu.
Export results
R / JamoviExport statistical tables, significance letters, model results and paper figures.
Save both the analysis script and the final results table so you can recompute during revisions.
Articles & Guides
- Don't want to buy SPSS? How to choose between JASP, jamovi and R
Compares JASP, jamovi and R on one public data set, gives a runnable R workflow, and uses real output to show which settings to check when software disagrees.