RESEARCH METHODS : PROBABLITIY AND SIGNIFICANCE Flashcards
each study has how many hypothesis
2
what is an orgianl or alternate hypothesis
states there will be an effect
whats the symbol for original or alternate hypothesis
H1
What is a null hypothesis
states there wont be an effcet
nul hypothesis symbol
H0
GIVE EXAMPLE OF ALTERNATE HYPOTHESIS
there will be an effect n pt depression score
give example of null hypothesis
states there wont be an effect
what do stat test tell us about null hypothesis
whether or not to reject null hypothesis
if we reject the nul hypothesis whta does this say about the study
good and stat test told me results are good
we look for what rather than truth
significance
lets say we find a difference and everyone imporved this dont means the results are …….
because resuts could easily be down to
of value/significnat
chance
all studies ewmploy a significance level in order to
check for significant differences or relationships
significance levl is the point at which
point at wihc researcher can claim to have discovered a large enough difference or correlation within the data to calim an effcet has been found
signficnace level in short
point at whihc researcher can reject teh null hypothesis and accept teh alternative hypothesis
stat tests tell us whether
reuslts are significant o rnot
what is the accepted level of proability and waht you use unless told otherwise
0.05 / 5%
the usual level of signicfane is 0.05 how do you write this properly
p<_0.05
0.05 is the elvel at whihc waht is accepte
alternative/original hypothesis is accepted
what doe a 0.05 level of probaility tell us
less than 5% proabilty result occured by chance
if results come back as significant after conducting a stat test what does this mean probability wise
95% proability my result occured due t the manipulation of the IV on the DV ?
p of 0.05 measn even if researcher claims to have found asignificant difffrence/correlation
there is still up to a 5% chance that it isnt true for the target population in whihc the sample was drawn
why cna psycholoigsts never be 100% certain about a particular result
so they have
theyve not ested all members of the population under all possible circumnstances - this is not practical economically due to time and costs
so theyve settled on a conventional level of proability where tehyve prepared to accept results may have occured by chance
some studies have a 0.001 level of probaility this depdns on
what the researchers are trna achieve
5% is an accpetable
margin of error
choosing the wrong significance level can lead to
type 1 error
type 2 error
what is a type 1 error
when you choose a significance level thats too high such as 0.1/10%
the incorrect rejection of a true null hypothesis
flase positive
what happens when you have a type 1 error
reult in rejecting a null hypothesis thats true
fasle positive
null hyp rue
alt hyp wrong
say results are signficant when tehyre actually not
what is a tyoe 2 error
the failure to reject a false null hypothesis
a false negative
when you choose a significance level that is too low e.g 1% or 0.01%
what happens when you have a type 2 error
null hypothesis is accepted but alt hypothesis should have been accepted
because its true
why dopsychologists facour hte 5% level of significance
best balances the risk of making a type 1 or type 2 error
once a stat test been calculated whats teh result
a number - calculated value
to check for stat significance teh calculated value must be
compared with a critical value
what is a critical value
a number that tells us whether or not we can reject the nulll hypothesis and accept eh alternative hypotheis
each stat test has its own
table of critical values developed by staticians
what is the rule of R
Stat test with the letter R in their name re those where calculated value must be = to or more than critical vlue
+ there is an r in more
how do we know which critical value ot use
one or two tailed test
numbe of participants in the study /degrees of freedom
level of significance
when do we use a one tailed test
directional hypothesis
when do we use a two tailed test
non directional hypothesis
when and why do probailty levels double
when two tailed test is being used
as they are a more conservative prediciton
number of pt in the study usually appear as
n value on the table
when would a stringetn / low level of singifanec be used e.g 0.01
studies where there may be a human costs such as drug trials
or one off studises where for practical reasons cannot be reoeated in teh future
what happens if there is a large difference between the calcualted and critical values
in teh preferred direction
the researcher will check more stringent levels as the lower the p value
the more statistically significnant the result