Interpreting data

[#1]

So, i have watched video after video to try and interpret this data. I surveyed the black river watershed averaging the riparian buffer widths of 13 sites. I am trying to interpret how this might affect sediment size (D50 ) and percent of intolerant at these transects. I know this has no statistically significant results. But my professor said I need to say more then that. I have also done spearman's r, kendalls, graphs and correlation probabilities, I just don't really know how to interpret all of this. There is a slight negative trend in the data. Can anyone help me? 

Variable	by Variable	Spearman ρ	Prob>|ρ|
D50	Rip Buf Avg	-0.3054	0.3103
% Intolerant	Rip Buf Avg	-0.2418	0.4262
% Intolerant	D50	-0.2724	0.3680

[#2]

Hi @tgrandjean1 ,

  Not knowing a whole lot about the specifics of what you're studying, it's hard to provide much in the way of specific feedback. That being said:

1. Clearly the Rip Buf Avg has some kind of (if small) negative correlation with D50 and D50 has some (small) negative correlation with %intolerant, but %intolerant and Rip Buf Avg have essentially 0 correlation.
2. Since the response is D50, do you have a model for RBA and %int? For example, have you tried fitting a model to D50 where the factors are RBA, %int and RBA*%int - that is with RBA crossed with %int? Without a model or some concept to connect the response with factors, it'll be hard to say a whole lot more than there's no statistical significance. In addition to this, it seems that you should also have a column that records which site the particular measurement was taken at, as this could be a contributing factor in the model.
3. If you are really looking only at linear correlations you'll want to do a Fit Y by X and then fit a line and look at some of the statistical tests that it does, like lack of fit, or what the r^2 value is. Additionally, you'll want to consider what the y-intercept is and think about whether or not it is a meaningful number when the X factor is 0.
4. Have you asked for clarification from your prof as to why they are saying you must say more than that? Sometimes being concise is better.
5. Can you share your data table? You can always anonymize and standardize it. This could help others to provide more meaningful feedback to you without knowing the full details of the data.
6. Are there multiple measurements per site? It could be that the site location plays a bigger role than the RBA or %int.

Hope this helps,

DS