A bar graph is the best choice for showing the means of several different treatments.

Why a bar graph works well

  • Each treatment becomes a separate bar on the x‑axis, making comparisons very clear.
  • The height of each bar represents the mean for that treatment, which people can visually compare quickly.
  • You can easily add error bars (for standard deviation, standard error, or confidence intervals) on top of each bar to show variability and uncertainty.

How you might set it up

  • Put treatment names or labels on the horizontal axis.
  • Put the response variable (e.g., mean score, mean growth, mean response time) on the vertical axis.
  • Use one bar per treatment (or clustered bars if you also have groups like “male/female” or “before/after”).
  • Add error bars for each mean if you want to communicate reliability of the estimates.

Brief comparison with other graphs

  • Line graphs are better for trends over time or ordered levels, not for separate, unordered treatments.
  • Pie charts are for parts of a whole (percentages), not for mean comparisons.
  • Scatter plots are for relationships between two continuous variables, not for comparing several group means.

So, if your question is “what type of graph would be best for showing means for several different treatments?” , the standard and most effective answer is: a bar graph (with error bars, if possible).

TL;DR: Use a bar graph with one bar per treatment, and add error bars if you want to show variability.

Information gathered from public forums or data available on the internet and portrayed here.