statistical tests

Created by

vaishali bansal

Cards (22)

what do statistical tests tell us
they tell us about probability - so the answer is a probability value
what is the purpose of stats
if the probability indicates that there is a statistically significant difference between the IV conditions , then the alternative hypothesis is true
why do we have statistical tests
to decide if there is something meaningful in our data
whether our results were a fluke
what are the different statistical tests
sign test
chi-squared
wilcoxen
mann whitney u
spearman's rho
unrelated t-test
related t-test
pearsons
what is the difference between a test of difference and a test of association
a test of difference looks for a difference between two samples or two conditions
a test of correlation looks for a relationship between two variables
what are the three types of levels of measurements
nominal data
ordinal data
interval data
what is nominal data
the data produced is just frequency scores- the number of people in a category
what is ordinal data
a measurement scale where the scores are ranked- the intervals between each integer isn't standardised
what is interval data
a measurement of data on a continuous scale of equal intervals - if it has absolute zero then it is interval data
if there isn't an absolute zero then it is ratio data
what is the difference between unrelated and related design
related - matched pairs and repeated measures
unelated - independent groups
what is a parametric test
when the test makes an underlying assumption of data
which tests are parametric
unrelated t-test , related t-test and pearsons - all use interval data
what do we need to know in order to interpret the outcome of different statistical tests
levels of significance
observed /calculated values
critical values
degrees of freedom
what are 'levels of significance'
the accepted level of significance in psychology is 5% - this means there is less than or equal to a 5% chance that the null hypothesis is true - we can be 95% confident that we have made the right decision to reject the null hypothesis and accept the alternative hypothesis - this is written as p≤ 0.05
what is the 'observed value'
you do a statistical test and a number is produced at the end of that test - this is the observed / calculated value
what is a critical value
each statistical test has tables of critical values - this tells us the cut-oof point at which the results are significant
what are 'degrees of freedom'
the number of 'observations' in the data that are free to vary when estimating statistical parameters
what are the two sources that may result in a sample being different from the target population
sampling error - an error which has occurred due to chance due to random error
sampling bias - a constant error due to inadequate design
what is a type 1 error
when the null hypothesis is incorrectly rejected and the alternative is incorrectly accepted
a false positive
type 1 error is most serious mistake to make
likelihood of making the error is equal to the level of significance
as the significance increase so does the likelihood of making a type 1 error
what is a type 2 error
when the null hypothesis is incorrectly accepted and the alternative is incorrectly rejected
a false negative
less serious
likelihood of making a type 2 error is equal to the level of significance
likelihood of this increases as the level of significance decreases
why is the 5% level of significance most commonly used in psychology research
it strikes a balance between the risk of making type 1 and type 2 errors
what are the steps involved in doing the sign test
calculate the sign of difference
calculate the total number of positive and negative signs
the calculated value is the less frequent sign
find n- this is the number of participants excluding any equal scores
establish if the hypothesis is directional or non-directional
find the critical values from a table of values
compare the calculated value with the critical value to see if there is any significance - o ≤ c