The Student Room Group
Uni students - Complete our survey >>
Uni students - Complete our survey >>
Uni students - Complete our survey >>

+/- sign.

I've never had to deal with this sign before, what exactly does it mean?


Cohen (1988) provides criteria for assessing the strength of r (also known as effect size): Small: r +/- 0.1 Medium: r +/- 0.3 Large: r +/- 0.5What does the +/- mean in this context?
It's called the plus or minus sign.

The correlation coefficient can also be used when the data are binary. R can vary in magnitude from −1 to 1, with −1 indicating a perfect negative linear relation, 1 indicating a perfect positive linear relation, and 0 indicating no linear relation between two variables.
Original post by natalie427
I've never had to deal with this sign before, what exactly does it mean?


Cohen (1988) provides criteria for assessing the strength of r (also known as effect size): Small: r +/- 0.1 Medium: r +/- 0.3 Large: r +/- 0.5What does the +/- mean in this context?


Depends on the context. In Mathematics a ±\pm sign gives strictly 2 different values, however in other applications it can give a range of values.

For example 3±0.53\pm 0.5 can provide a range between 2.5 and 3.5, or give you only strict values of 2.5 and 3.5.
Reply 3
Avatar for natalie427
natalie427
OP
12
Original post by Kvothe the Arcane
It's called the plus or minus sign.

The correlation coefficient can also be used when the data are binary. R can vary in magnitude from −1 to 1, with −1 indicating a perfect negative linear relation, 1 indicating a perfect positive linear relation, and 0 indicating no linear relation between two variables.


I understand all that a but this is referring to the strength of the correlation, I just can't work out what that sign suggests.
Reply 4
Avatar for natalie427
natalie427
OP
12
Original post by RDKGames
Depends on the context. In Mathematics a ±\pm sign gives strictly 2 different values, however in other applications it can give a range of values.

For example 3±0.53\pm 0.5 can provide a range between 2.5 and 3.5, or give you only strict values of 2.5 and 3.5.


Actually Ive just worked it out

Small = +/1 0.1


Medium = +/- 0.3

Large = +/- 0.5


So for example, 0.01 would be small
0.2 would be a small/medium
0.3 would be medium
0.4 would be medium/large
0.5 would be large
Original post by natalie427
Actually Ive just worked it out

Small = +/1 0.1


Medium = +/- 0.3

Large = +/- 0.5


So for example, 0.01 would be small
0.2 would be a small/medium
0.3 would be medium
0.4 would be medium/large
0.5 would be large


Yes in your context it can mean that whatever value you take, like 0.3, then the affect size will be medium regardless whether it is negative or positive.
Reply 6
Avatar for natalie427
natalie427
OP
12
Original post by RDKGames
Yes in your context it can mean that whatever value you take, like 0.3, then the affect size will be medium regardless whether it is negative or positive.



Yes the effect sizes of the correlation is independent to whether the correlation is positive or negative.

I just was unsure of the sign, the correlation and effect size I'm aware of .

Quick Reply

Related discussions

Latest

Trending

Trending

Articles for you