All language subtitles for 05 - Calculate the correlation between two columns of data

- [Instructor] One way to analyze

the relationship between two sets of values

is to calculate their correlation,

which is a measure of how much

the two sets of values tend to move together,

whether in the same direction,

different directions, or not at all.

My sample file is 05_05_SingleCorrelation

and you can find it in the Chapter Five folder

of the Exercise Files collection.

I have two columns of data

and you can see them there in worksheet columns A and B,

and then in cell D3, which I have selected,

I can calculate the correlation

between those two sets of values.

So in cell D3, I'll type an equal sign

and the function I want to use is coral,

or correl, C-O-R-R-E-L,

finds the correlation coefficient between two data sets.

I'll press tab to select it.

And now, I just need to identify the data.

So I will select cells A3 through A12,

then type a comma, do the same thing for B3 and B12.

And these two arrays do need to be the same size.

They need to have the same number of values.

Right parentheses and enter.

And I see that the correlation is 0.69, so almost 0.7.

That is a fairly high value.

But what I'd like to do is look up

to see whether this is a significant value.

So I will need to look to see

if this correlation value of 0.69 is significant

by looking at a table.

I have displayed the correlation lookup table

for two-tailed test, in other words,

we can have a positive or negative difference,

that I showed in the previous movie.

I have 10 samples, so I will look in the row

that starts with 10 in the N column,

and I can see how significant my result is.

Remember that my correlation coefficient

was 0.6948, so 0.69.

And if I look in the 10 row,

I see that we are significant at the 90% level, that's 0.1,

and that we are significant at the 95% level, that's 0.05.

However, if I look to the right in the 98% column 0.02,

I see a comparison value of 0.71,

so our correlation is not significant at that level.

However, 95% meaning that we have a 95,

or 19 out of 20 chance that the result is significant

is a pretty good result and indicates there is probably,

probably a relationship between those two sets of values.