How would you determine if a dataset follows a normal distribution?

In my final year at college, I worked on a project analyzing survey data to understand student satisfaction. To check if our dataset followed a normal distribution, I created a histogram and a Q-Q plot. The histogram showed a bell-shaped curve, and the Q-Q plot indicated that our data points closely followed the diagonal line. Additionally, I conducted a Shapiro-Wilk test, which resulted in a p-value above 0.05, suggesting that our dataset was normally distributed. This analysis helped us confidently apply statistical tests that required normality, such as t-tests, to draw conclusions about student satisfaction levels.

While volunteering at a local non-profit, I helped organize community events and collected attendee feedback. To determine if the feedback scores followed a normal distribution, I first plotted the scores in a histogram, which showed a roughly symmetric shape. I also calculated the skewness and kurtosis, finding that both values were close to zero, indicating a normal-like distribution. This understanding allowed us to apply statistical analyses to improve future events effectively. By ensuring our data met normality assumptions, we could confidently use parametric tests to assess satisfaction levels and plan more successful gatherings.

In my first job as a data analyst, I often encountered datasets where normality needed to be assessed. For instance, when analyzing customer purchase data, I used a combination of visual tools like histograms and box plots to visually inspect the data's shape. I also applied the Kolmogorov-Smirnov test to statistically evaluate normality. Understanding whether our data was normally distributed was crucial since many predictive modeling techniques assumed this. By effectively communicating our findings, I could guide our team in choosing the right analytical methods, ensuring our strategies were based on solid statistical ground.