Data handling analysis

Question 1

Inferential tests

Answer 1

Tests of significance that tells us the probability that the results of the analysis could have occurred by chance

Question 2

Significance

Answer 2

When results of analysis are significant it means that it is possible to say to some degree that they are not due to chance

Question 3

Probability and significance

Answer 3

p < 0.5 the probability that the results occurred by chance is less that 0.5. This means the result is significant.

p > 0.5 the probability that the results occurred by chance is greater that 0.5. This means the result is insignificant.

p = 0.5 the probability that the results occurred by chance is equal to 0.5. This means that the result is significant.

If there is significant results then the hypothesis can be accepted and the null hypothesis rejected.

If the results are insignificant then the hypothesis can be rejected and the null hypothesis accepted.

Question 4

Type 1 and type 2 error

Answer 4

It is possible to make two kinds of mistakes when interpreting statistical data.

Type 1 error: rejecting null hypothesis and accepting hypothesis even though findings are insignificant

Type 2 error: retaining the null hypothesis even though the hypothesis is correct (false negative).

Question 5

Levels of measurement

Answer 5

Levels of measurement refers to the relationship among values assigned to variables.

Question 6

Nominal level of measurement

Answer 6

Data that can be put into categories. Can be presented in bar graph.
The most basic level of measurement where data is simply categorised. Gives little information and often just tells how many people are in a group.

mode

Question 7

Ordinal level of measurement

Answer 7

Continuous data.
Can be presented in histogram.
Give more info than nominal scales
Ordered relationship between variables observed (rank)
Usually based on opinion so more subjective than objective.

median

Question 8

Interval level of measurement

Answer 8

Continuous data. Represents things set in stone like time, weight, length.
Present more info than ordinal and nominal
The most complex level of measurement where the difference between two values is meaningful.
There is an equal gap between each unit of measurement.
Zero doesn’t mean zero

mean

Question 9

Parametric tests

Answer 9

They make assumptions that:
Population drawn must be normally distributed
Variance of data should be approximately equal
Should have at least interval data
Should be no extreme scores

Includes:
Related t test
Unrelated t test
Pearson r

Question 10

Non- parametric tests

Answer 10

Mann-Whitney
Wilcoxon 
Chi Squared 
Sign test 
Spearman's Rho