# Determining if linear regression is appropriate

I am designing an experiment, and I want to know if using a hypothesis test on the linear regression model to analyze my data is appropriate. The experiment will have five independent variable values (continuous variable), with five trials for each, for a sample size of 25. I'm not sure if using linear regression is appropriate since the data is not really a scatterplot; visually, it's five points above each of five values on the independent variable axis. Would this test be appropriate, or is there something better I can use?

Edit: One idea I had was averaging the result for each independent variable value, but would this impact my regression statistics in a way that the sample is treated as having size 5 rather than 25?

• Please can you describe the experimental design and let us know your research question(s). – Robert Long Sep 24 '20 at 6:41
• In the experiment, I will water plants with different concentrations of salt water. The independent variable is salt concentration and the dependent variable is height grown. The research question is what is the effect of the concentration of a particular salt on the growth of a particular plant. There will be five salt concentrations and I will grow five plants at each of those concentrations. – physicsaficionado Sep 24 '20 at 12:40
• How many times do you measure the plants ? – Robert Long Sep 24 '20 at 13:34
• 5 times over a 15-day period. – physicsaficionado Sep 24 '20 at 17:24
• So are you interested in the rate of growth, or just the final measurement, or something else ? – Robert Long Sep 24 '20 at 17:26