statistical terminology

Question 1

null hypothesis

Answer 1

no significant difference between specified populations that we are looking at. For bivariate data we are saying there is no correlation. For chi-squared tests we are saying there is no dependence.

Question 2

alternative hypothesis

Answer 2

There is a difference between populations. Our statistical tests are seeing if this hypothesis is possible.

Question 3

When do we use a 1 or 2 tail test

Answer 3

1 tail test = if we were absolutely certain that a negative correlation was not possible we can assign 100% of our critical region to just option 1. I.e there will be a positive correlation or no correlation at all.

2 tail test = We assign each possibility 50% of our critical region

Question 4

Types of sampling

Answer 4

• random
• stratified
• cluster
• quota
• oppurtunity
• volunteer

Question 5

When is a chi squared(x2) test used

Answer 5

used to investigate whether distributions of two categorical variables differ from one another

Our null Hypothesis in any chi-squared test is..
“variable A is independent of variable B”

Our alternative hypothesis is
“variable A is dependent on variable B”

Question 6

How to work out product moment correlation

Answer 6

Enter stats mode
Choose regression (a +bx)
Enter the two sets of data values
Get the regression value

Question 7

What do the results of product moment correlation show

Answer 7

-1 (perfect negative correlation) to +1 (perfect positive correlation).

Question 8

How to work out spearmans rank correlation

Answer 8

1. rank the data set
2. calculate the difference between the rank scores then square them
3. add up all of the differences squared(d2) and count how many numbers there is(n), then use the formula

1-
6xd2
/n(n2 - 1)