The Stats Test - Probability and Significance Flashcards
Three types of data
Ordinal
Nominal
Interval
Ordinal data
Data that is in ranked order
Example of ordinal data
First to last
What measure of central tendency does ordinal data use?
Median
Nominal data
Data in frequencies that is put into categories
What measure of central tendency does nominal data use?
The mode
Interval data
Data that is in a fixed unit of measurement
Is interval data a subjective or objective way of gathering information? Why?
Objective because it isn’t based on interpretation
Example of interval data
KG or temperature
What measure of central tendency does interval data use?
Mean
Strength of the mean
Uses every value in the data
Limitation of the mean
Vulnerable to outliers affected by extreme scores
Strength of the median
Least affected by extreme scores
Limitations of the median
Doesn’t use every value in the data
Strength of the mode
Not affected by extreme scores
Limitation of the mode
Not based on all values
What does the range tell us?
The difference between the smallest value and the largest value
What does standard deviation tell us?
Tells us how far the spread of data/scores are from the mean
What will the values be if the SD is 3?
The values will be 3 above and 3 below the mean
What do we use statistical testing in psychology?
To see if the data is statistically significant or not and whether out hypothesis is supported or rejected
How can the likelihood of the results having occurred by chance be found?
By carrying out the test multiple times to see if the results are statistically meaningful
What will out hypothesis be if our findings are statistically significant? What will this show?
Our hypothesis will be supported, showing us that the study is predictable
How were Milgram’s results statistically significant?
Because 100% of the participants obeyed to 300V and 65% obeyed up to 450V, allowing the conclusion that we will obey orders of an authority figure even if we are unsure about what we are told to do
Probability in psychology (P) =
P = < or equal to 0.05
What test is used when a hypothesis predicts a difference between co-variables or association?
Chi Square Test
Anagram for Chi Square Test
CANDI
C of CANDI
Chi Square
A of CANDI
Association
N of CANDI
Nominal
D of CANDI
Difference
I of CANDI
Independent groups
CANDI
Chi square
Association
Nominal
Difference
Independent groups
What data does the Chi Square Test deal with?
Nominal data
Formula for degrees of freedom
(Number of columns - 1) x (number of rows - 1)
Symbol for Chi Square
X squared
Unrelated t test anagram
UNIITD
U and N of UNIITD
Unrelated
I of UNIITD
Independent groups
I of UNIITD
Interval data
T of UNIITD
Time
D of UNIITD
Difference
UNITD
Unrelated
Independent groups
Interval data
Time
Difference
Related t test anagram
DIRRT
D of DIRRT
Difference
I of DIRRT
Interval
First R of DIRRT
Repeated measures
Second R and T of DIRRT
Related T
DIRRT
Difference
Interval
Repeated measures
Related T
What is the MIDO anagram used for?
The Mann-Whitney experiment
M of MIDO
Mann-Whitney
I of MIDO
Independent
D of MIDO
Differences
O of MIDO
Ordinal
MIDO
Mann-Whitney
Independent
Differences
Ordinal
Way of remembering MIDO
MIDO is the “psychology dog”
Anagram used for the Wilcoxon test
WORD
W of WORD
Wilcoxson
O of WORD
Ordinal
R of WORD
Repeated measures
D of WORD
Difference
WORD
Wilcoxson
Ordinal
Repeated measures
Difference
Anagram for Pearson’s r
CIP
C of CIP
Correlation
I of CIP
Interval
P of CIP
Pearson
CIP
Correlation
Interval
Pearson
If you’re trying to find the correlation of two variables where one is the ordinal level of data and one is the interval level of data, what test do you use?
Spearman’s Rho
In psychology what is the default p value?
Default probability value is 0.05
What does the 0.05 default p value mean?
Means you are 95% sure that your results were not due to chance
The lower the p value, the more…
Confident you are about the results not being due to chance
When studying humans, why isn’t the p value 0.01?
Because you cannot predict what every human’s behaviour will be like due to factors such as individual differences, free will and human error
Why do we use statistical tests in psychology?
To establish whether or not the results of an experiment were due to chance
Stringent
Strict or precise
The more stringent the level of significance is, the more…
Challenging it is to get your results to show that your hypothesis is supported
The less stringent the level of significance is, the less…
Challenging it is to get your results to show that your hypothesis is supported
In what scenario would a high p value be suitable?
Drug testing because you need to ensure the drug works before giving it to patients
What does using a 0.05 level of significance minimise?
Minimises the chance of making a type 1 or 2 error
Type 1 error
When the p value is very large there is a danger that we will take a chance result and decide it’s significant when it’s not
What is type 1 error also known as?
False positive
What is a chance result?
A result not representative of other results
What does type 1 error make easier to do?
Makes it easier to get the results looking like they were statistically significant when there is a bigger margin of error
Type 2 error
When the p value is very small there is a danger that we will take a significant result and decide that it was caused by chance
What is type 2 error also known as?
False negative
When is a type 2 error likely to occur?
When the level of significance is too stringent, making it harder to provide support for the hypothesis, meaning you could potentially reject a valid hypothesis
