Statistics Flashcards
what issue do stats mainly serve to solve?
to workout the likelihood that the results were obtain by chance
what shape is the probability distribution
bell curve
what does the null hypothesis help to determine?
what does it assume?
how unlikely our empirical result was under the chance distribution
null hypothesis assumes that result was all due to chance
what can we use to represent the chance distribution?
t-distribution and z-distribution
when can we only use z distribution?
when we know the population standard deviation
what is used when you do not know the population standard deviation?
t-distribution
what do you calculate the empirical t value
you take the difference between the observed mean, and the population mean, and divide it by the standard error of the mean
what is the population mean under the null hypothesis ?
zero
what changes the shape of the t distribution?
the degrees of freedom
smaller degrees of freedom change the shape of the t distribution by in what way?
the smaller the df, the “fatter” the tails will be, meaning more results cam call under this
….
the the observed t value, given the null hypothesis that t=0
what does a one sample design involve ?
comparing group data to known values (like IQ or “an hour” waiting time)
what is another name for a between groups design ?
an independent measures design
independent samples t test is used for…?
independent measures design
what is the main benefit of using an independent measures design
there will be no learning effects due to repeated exposure
what is the main 2 disadvantages of independent measures design?
- people in different groups might greatly vary
- we cannot study how things can change overtime
what is another name for the within group design?
repeated measure design
what are 2 main advantages of repeated measures designs?
baseline factors like personality etc don’t have to be accounted for as they will affect both conditions equally
- can study changes in behaviour over time
what are 2 main disadvantages of repeated measures designs?
- measurements are not independent
- people cannot be naive in second round
- need to factor in order effects (who does what condition first or second)
what is the one sample t test formula?
M - u
t = ________
S subscript M
what is the formula to work out the standard error of the mean?
S
S (subscript) m = _______
square root
of n
the standard deviation is what? how is it represented in formulas?
the square root of the variance
S
how do we get the variance number?
the sum of squared differences divided by degrees of freedom
why do we square the final value?
because we care about the magnitude not whether or not its above or below the mean
t or f. the t value also quantifies the effect and tells us how strong the effect is
false
what is the standard deviation formula?
SS
s^2= _____
df
then square root this number
what is the t equation for the independent measures t test?
(Ma-Mb)-(ua-ub)
t= _____________
S subscript
(Ma-Mb)
what is S
(Ma-Mb)
the difference between two means
standard error of the mean difference
what is a key component of working out the standard error of the mean difference?
the pooled variance
how do you calculate the pooled variance?
how does this look in a formula?
summing up the summed squared differences (SS) from each condition and dividing it by the sum of the df
SSa + SSb
s2p = ___________
df a + df b
what is the short hand for pooled variance?
2
S
p
the pooled variance is just what?
the average of the two sample variances
what mean do we need to acquire in a paired samples t test ?
the difference mean
so we subtract the result from condition b from the result of condition a for all participants, and then find the mean of all these differences
true or false, cohen’s d gives an estimate of effect size that is independent of sample size
true
what does the d refer to in cohen’s d?
the mean difference divided by the standard deviation
what are the conventional cut offs varying sizes in cohen’s d result?
small: 0.2-0.5
medium: 0.5 - 0.8
large: greater than 0.8
what does r^squared mean?
variance explained
True or false, r^2 is dependent on the sample size
true
what is the formula for r^2
t^2
r^2= _______
t^2 +df
what are the conventional cut offs of varying sizes in r^2
small: 0.01 - 0.09
medium: 0.09 - 0.25
large: bigger than 0.25
what is a confidence interval?
setting parameters based on the assumption that the sample mean should be relatively near to the corresponding population parameter
