# What statistical evidences are needed to prove/disapprove a regression

I will be very honest, This is related to an assignment but i am not asking for the answers. I am just asking for guidance since it is kind of throwing me off.

Background The question that i am working on has 3 parts. The question refer to a scatterplot which has a regression line. The question says that "without using any" other information from the article, how would you say that X is impacting Y. I used the concept of slope in straight line and laid down the argument that since the slop is positive, it means that x is impacting y bla bla. I was able to answer the second question as well which was asking that if you have to go against it, how would you go. Again, just looking at the image.

Now the third question is asking that if you need any other statistical evidence, what it would be. I am not sure how to answer this. I mean the question is so vague that i don't even know what to google or read in different docs. I have read multple online tutorials but still confuse.

Can anyone give me a pointer? thanks

• It would help if you posted the verbatim question. (There’s nothing wrong with asking for help with a confusing homework question.) As it stands, the question is far too vague to give a decent answer, though part of the goal of the question might be to be vague and let you be creative as a scientist with your response. – Dave Oct 25 '20 at 1:56
• I actually posted the 3rd part (re-worded). The case study talks about obesity and lactose intolerance in a specific (region) population. The questions is asking to interpret the conclusion just from the figure and not from the article (the goal is to make us understand how to interpret graphs?). thats why i am getting confuse to that what statistical evidence can we add from a scatterplot point of view. – Em Ae Oct 25 '20 at 2:45
• Then maybe show us the scatterplot? – kjetil b halvorsen Oct 25 '20 at 13:14