Here's a breakdown:
Acceptable:
* Clarity: The most important factor is that the graph clearly communicates the relationship between variables. If using 3 units allows for a clear representation of the data, then it's acceptable.
* Data Range: If the data range is such that 3 units per division makes the graph a reasonable size, then it's fine. For example, if you're plotting data from 0 to 12, using a scale of 3 units per division would result in a graph with 4 divisions.
* Comparison: If you are comparing multiple graphs, using a consistent scale (even if it's 3 units) can make comparisons easier.
Not Acceptable:
* Misleading: If a scale of 3 units distorts the data or makes it difficult to see trends, then it's not acceptable. For example, if your data points are all clustered very closely together, using a scale of 3 units might make it look like there's no change.
* Unnecessary Complexity: While using 3 units might not be inherently bad, it could make the graph unnecessarily complex. A scale of 1, 2, 5, or 10 units might be easier to read and interpret.
General Recommendations:
* Use Standard Scales: Scales of 1, 2, 5, or 10 units are generally preferred because they're easy to read and work with.
* Choose an Appropriate Scale: Choose a scale that allows you to clearly display the data range and any trends.
* Be Consistent: Use the same scale for both axes if possible.
Ultimately, the best scale to use depends on the specific data and what you're trying to communicate.
