Research Methods - Y13

Question 1

What are inferential statistics?

Answer 1

Allow us to infer the significance of the results from the data
see how likely they are to be due to chance

Question 2

What is a type 1 error?

Answer 2

False positive
Hypothesis wrongly accepted
(Rejects null)

Likely when probability value too high (eg 0.1)

Question 3

What is a type 2 error?

Answer 3

False negative
Wrongly reject hypothesis
(Accept null)

Likely when probability value too low (eg 0.01)

Question 4

Probability

Answer 4

Likelihood that any difference or association between groups has occurred simply due to chance

Question 5

what is a significance level?

Answer 5

used to establish probability of results being due to change
number between 0 and 1

smaller significance level, eg 0 = less likely to be due to chance
usually use 0.05 (5% likelihood due to chance)

Question 6

what is a critical value?

Answer 6

provided in a critical value tables

compared with observed value (value calculated by inferential tests)
to see if results are significant

Question 7

one tailed test

Answer 7

directional
researcher has predicted outcome and stated the direction of the results

Question 8

two tailed test

Answer 8

non-directional
researcher has predicted an outcome but hasn’t stated direction of results

Question 9

what is N?

Answer 9

N = number of values in the data set (eg number of participants)

Question 10

nominal data

Answer 10

frequency count for distinct categories
Data only in one catagory
(No order)

Question 11

ordinal data

Answer 11

measurements can be placed on a scale, in rank order
units are not the same size

eg aggression on a scale of 1 to 10

Question 12

interval data

Answer 12

data is on a scale
where each unit is the same size

eg length in cm

Question 13

how to decide which test to use (and table)

Answer 13

determined by:
- experimental design
- research aim - difference (2 separate conditions) or correlation (finding a relationship between variables)
- type of data (level of measurement)

table:
cramming sessions make weekends stressful, use regular preparation

DIFFERNECE CORRELATION
UR R
chi2 sign
mann w wilcoxon spearmans
unr. t rel. t pearsons

Question 14

how to do sign test

Answer 14

used on related designs

1. difference between participants scores calculated
   (+ if score increased, - if decreased)
2. add up number of + and -
   smallest one is observed value
   (eg 5- and 2+ then observed = 2)
3. compare observed to critical
   observed must be SMALLER or = to critical = significant

Question 15

when does the observed need to be GREATER than or equal to critical?
to be significant

Answer 15

released t-test
unrelated t-test
chi squared
spearmans
Pearsons

Question 16

when does observed need to be LESS than or equal to critical?
to be significant

Answer 16

mann-whitney
sign
wilcoxon

Question 17

why do we use 5% level?

Answer 17

Means making a type 1 or 2 error less likely

Large enough to mean results are significant
But not too large to mean that insignificant results are published