Write the predicted direction before you look. Then compare two groups from a CSV: simple versus choice reaction time, or two inventory subscales. Report two means, two ns, and whether the difference matched the prediction.
Write your prediction first, then load data.