Login | Register

Welcome, Guest. Please login or register.

August 12, 2020, 10:58:03 am

Author Topic: Research investigation - interpreting  (Read 59 times)

0 Members and 1 Guest are viewing this topic.

Rakuu

  • Adventurer
  • *
  • Posts: 13
  • Respect: +1
Research investigation - interpreting
« on: August 01, 2020, 07:21:00 pm »
0
Hi guys,
Need a little guidance in interpreting my data. What exactly is p and R value for (in the dumbest way)?
My other data also happens to have a p value but no R value. Did some research about it being related to null hypothesis but I'm not sure if I fully get it.

 My collected datas are usually under p<0.05 - meaning it's against null hypothesis and then what ?? Others are about p<0.001. The author didn't really talk about it, he was just stating it. Is this relevant enough for me to talk about or I could just state it as well (eg. provide no context on why the values are like that).

Apologies if this sounded dumb but I'm a bit lost, most data I usually collect only has to do with r^2 values.  :) *hopefully I attached a photo of it since I dont know how to*
« Last Edit: August 01, 2020, 07:22:49 pm by Rakuu »

Bri MT

  • VIC MVP - 2018
  • Administrator
  • ATAR Notes Legend
  • *****
  • Posts: 3849
  • invest in wellbeing so it can invest in you
  • Respect: +2862
Re: Research investigation - interpreting
« Reply #1 on: August 02, 2020, 09:27:17 am »
+2
Not dumb at all!

The p value is the probability of the results occurring under the null hypothesis. At a high school level (and tbh lots of uni students think this way too rather than a more rigorous definition) the p value is the likelihood of you having those results just from chance when there is no actual relationship between DDT concentration and organic matter or pH.

If you're trying to say "yeah there IS a relationship" then you want it to be very unlikely that the only reason you see a relationship in the graph is because of chance. Specifically, we say that if there's less than a 5% chance of getting those results when there's no relationship (p < 0.05) then those results are significant. The smaller that chance the more likely there is an actual relationship so p < 0.001 (less than a  0.1 % likelihood of getting those results if there's no relationship) is more strongly significant than p = 0.49 for example. Some people have strong views on whether a result closer to p = 0.05 should be talked abut differently than p < 0.001

You don't really need to talk about this, it's just there to say "yeah there very probably IS an actual relationship rather than just chance making my data look like that"

R tells you about how much the variables are associating/correlating. You can think about this as being like how close the data points are to actually fitting on that red line of best fit. You can also think about it as being "if I know what the pH is, how easy is it for me to accurately guess what DDT concentration will be?"

Together, R and p are telling you that:
- seeing this relationship is very unlikely to just be chance / the relationship probably does actually exist ( p value)
- the relationship is fairly weak ( R value)

It's very important for the author to have reported these things but you don't need to have a good understanding of them for QCE chem.


This is a topic that confuses a lot of people; don't feel bad if you're still not confident on it!

Please feel free to let me know if that explanation should be brought to a more basic level or if you have any more questions :)