The Student Room Group

a level biology evaluation question

in the mark scehme it says:"No statistical test, so do not know if differences are significant
OR
No statistical test, so differences could be due to chance;"

I don't understand what is meant by "no statistical test" and also how do we determine that there's no statistical test or difference is due to chance Screenshot 2023-04-23 202847.png
(edited 1 year ago)
Original post by hoyam
in the mark scehme it says:"No statistical test, so do not know if differences are significant
OR
No statistical test, so differences could be due to chance;"

I don't understand what is meant by "no statistical test" and also how do we determine that there's no statistical test or difference is due to chance Screenshot 2023-04-23 202847.png

Normally there would be "error bars" for each bar, however in these figures they are not present.

Spoiler


If the error bars overlap, there is no significant difference between the means (i.e. any difference observed is likely due to chance). If they do not overlap, the means are significantly different (i.e. any difference observed is likely not due to chance) :smile:
There are no error bars on the graph, which usually show the calculated standard deviation of the data (statistical test), so you can say that there is no statistical test as it is not shown on the graph.

If there were error bars, you would be able to determine whether there is/isn’t a significant difference, or whether it is/isn’t down to chance by looking at whether the bars overlap or not.
Reply 3
Original post by bl0bf1sh
Normally there would be "error bars" for each bar, however in these figures they are not present.

Spoiler


If the error bars overlap, there is no significant difference between the means (i.e. any difference observed is likely due to chance). If they do not overlap, the means are significantly different (i.e. any difference observed is likely not due to chance) :smile:

thank you :smile:
Reply 4
Original post by butterfliesss
There are no error bars on the graph, which usually show the calculated standard deviation of the data (statistical test), so you can say that there is no statistical test as it is not shown on the graph.

If there were error bars, you would be able to determine whether there is/isn’t a significant difference, or whether it is/isn’t down to chance by looking at whether the bars overlap or not.


thankss :smile:

Quick Reply

Latest

Trending

Trending