# Basics

Tutorials on the basics of statistics in the Excel environment using Real Statistics website and free downloadable software.

Topics:

### 8 Responses to Basics

hi
I have one variable which is function of four or five variables and I want to fined a relation between them using the statistic program
and I have a data between the dependent variable and the independent variables
so can you help me please ???

• Charles says:

Hi,
I am afraid that you need to give me more information. Are you looking for a correlation coefficient?
Charles

2. Joan says:

Hi Charles,
I have a problem which seems easy but I can’t quite resolve.
I have a survey which has categorical data, aiming to compare two groups where the general group has 56 respondents and the expert group has 6.
I tried to use a Chi squared test on SPSS but more than 20% of my cells have an expected frequency of less than 5. Somehow I can’t see the Fishers exact test results and I understand it’s because data was not in a 2×2 form.
What should I do? Chi square doesn’t work because the rules have been violated but there is no option for a Fishers exact test?

Joan

• Charles says:

Joan,
You say that the contingency table is not 2 x 2. What is the dimension? Fisher can be used with contingency tables other than 2 x 2.
Charles

3. Joan says:

The row is “experts vs non experts”, while the columns are “age range” (1 for age 30-34, 2 for 35-39, 3 for 40-44 and so on until more than 70).

• Charles says:

Joan,
I suggest that you use fewer age ranges. E.g. age 30-39, 40-49, 50-59, 60-69, 70 or more. In this case perhaps all cells will have values of 5 or more. With a 2 x 5 contingency table you can use Fisher’s Exact test (Real Statistics supports this).
Charles

• Joan says:

Thanks, Charles. I’ll try to regroup this. Alternatively would a Mann-Whitney work?

• Charles says:

Joan,
This depends on the nature of your data. Generally, Mann-Whitney is used to compare the medians of two groups — e.g. median age between men and women.
Charles