Inferential Stats

Question 1

What do Inferential Stats analyse?

Answer 1

• The probability of something occurring due to the influence of the IV, or whether it is due to chance.

Question 2

What does p <0.05 mean in Psychology?

Answer 2

• There’s a 5% chance of the results being due to chance and NOT the IV influencing the DV.

Question 3

What is Nominal Data?

Answer 3

• Categorical data

- Lowest Level of Measurement

Question 4

What is Ordinal Data?

Answer 4

• Ranked data, goes in a specific order.

Question 5

What is Interval Level Data?

Answer 5

• Real measurements that have equal intervals and mathematical meaning.

Question 6

What is Ratio Level Data?

Answer 6

• Same as interval, except there’s a true 0

- Age, weight etc.

Question 7

When do we use a Mann-Whitney U Test?

Answer 7

• When there is ORDINAL data and an IGD.

Question 8

When do we use the Wilcoxon Test?

Answer 8

• When there is ORDINAL data and an RMD, MPD.

Question 9

When do we use Spearman’s Rho?

Answer 9

• Testing correlations in ORDINAL data.

Question 10

When do we use Chi-Squared?

Answer 10

• When NOMINAL data is used in an IGD.

Question 11

What is a Type 1 error?

Answer 11

• When using a level of significance that is too high (e.g., p <0.10), you may reject a null hypothesis that is true.

Question 12

What is a Type 2 error?

Answer 12

• When a researcher rejects an experimental hypothesis that they should’ve accepted because they have been too harsh.

Question 13

Mann-Whitney U Process

Answer 13

1. Multiply na by nb.
2. Multiply na by na+1 and halve the result.
3. Add the results from 1 and 2 together
4. Add up all ranks for condition A
5. Subtract this from the previous sum –> Gives value of Ua.

Do the same to find the value of nb, but replace na with nb.

Observed Value = Lower of Ua and Ub. OV should be LOWER than CV to be significant.

Question 14

Wilcoxon Process

Answer 14

1. Calculate difference between 2 scores by taking 1 from the other.
2. Rank the differences, giving the smallest difference Rank 1. –> IGNORE any 0s.
3. Add up ranks for positive differences.
4. Add up ranks for negative differences.
5. T = figure that is the smallest when ranks are totalled (can be + or -)
6. N is no of scores left, IGNORE those with 0.
7. Compare T to CV - T must be LOWER than CV to be significant.

Question 15

Chi-Squared Process

Answer 15

1. Create a contingency table outlining the data and totals.
2. Calculate “expected frequency” (E) for each cell –> EF = Row total x column total / grand total.
3. Subtract the E from the Observed Value (O) for each cell –> O-E.
4. For each cell –> (O-E)^2
5. For each cell –> (O-E)^2 / E
6. Add all values from the previous step to get a final value of X^2.
7. Calculate DoF –> (rows -1) x (columns-1)
8. Compare to CV Table –> OV must be MORE than the CV.

Question 16

Spearman’s Rho Process

Answer 16

1. Rank scores for both variables from lowest to highest. –> 2 the same = same rank.
2. Calculate the difference (d) in ranks for each PP.
3. Square the difference (d) to produce D^2, and add up the column to get the total of D^2.
4. Find the no of PPs (N)
5. Multiply 6 to the answer in Step 3.
6. Square N and -1.
7. Multiply answer from step 6 by n to get n(n^2-1)
8. Divide the answer to step 5 by step 7
9. Subtract the answer to Step 8 from 1 to calculate the OV of X.
10. Compare OV to the CV to find if results significant. –> OV has to be MORE than CV to be significant.