# Comparing Number of Members

#### JackLee

##### New Member
Dear all,

Our organisation would like to compare the number of members between 2017 and 2018.
The aim is to see whether number of members has increased significantly or not.
The data will be transformed into monthly data (total of members by month).
What statistical technique can be use? Is 2 sample t-test appropriate?

Jack

#### Karabiner

##### TS Contributor
Without the transformation into monthly data, a Chi² test would have seemed useful (compare the 2 frequencies, with the Null hypothesis that frequency 2017 = frequency 2018). If you use 12 monthly frequencies instead, then you want to compare January 2017 with January 2018, February 2017 with February 2018 etc., I suppose? That would mean to use paired samples t-test rather than 2-sample t-test.

With kind regards

Karabiner

#### JackLee

##### New Member
Without the transformation into monthly data, a Chi² test would have seemed useful (compare the 2 frequencies, with the Null hypothesis that frequency 2017 = frequency 2018).
Did you mean Chi square Goodness of fit ??
(If the data were not transformed into monthly data, then there'd be only 2 figures)

If you use 12 monthly frequencies instead, then you want to compare January 2017 with January 2018, February 2017 with February 2018 etc., I suppose? That would mean to use paired samples t-test rather than 2-sample t-test.
Paired t-test?

#### JackLee

##### New Member
Yes.

Yes, n=12.

With kind regards

Karabiner
I had thought about using paired t-test earlier, but I thought it's not appropriate because I thought the 2 counts of
each month is not from the same person or thing.

#### Karabiner

##### TS Contributor
Its paired observations (2 observations for the same month of the year), therefore the according t-test.

With kind regards

Karabiner