# research methods ; probability + significance

PROBABILITY AND SIGNIFICANCE
GD
1 of 34
- the null hypothesis
dfgdg
2 of 34
what two things may hypotheses be?
directional / non-directional
3 of 34
what may these two types of hypotheses also be referred to as?
alternative hypothesis
4 of 34
why?
alternative to null hypothesis
5 of 34
what does null hypothesis state?
there is no difference between conditions
6 of 34
what determines which hypothesis is true + whether we accept or reject null?
stats test
7 of 34
- levels of significance and probability
dfgd
8 of 34
stats tests don't work on certainty but?
probability
9 of 34
all stats tests emply a significance level which is?
pt at which researcher can claim to have discovered significant difference / correlation within data
10 of 34
in other words pt @ which researcher cna do what?
reject null + accept alternative
11 of 34
what is the usual level of significance?
5% / 0.05 / p < 0.05
12 of 34
what does this mean?
probability that observed effect occured by chance is equal to / less than 5%
13 of 34
in effect when researcher claims to have found difference / correlation there's still 5% chance
it was a fluke
14 of 34
why can psychologists never b 100% certain about particular result?
they haven't tested all members of population under all possible circumstances
15 of 34
for this reason conventional probability level is?
5%
16 of 34
USE OF STATS TABLES
DFGDF
17 of 34
- critical value
dfdf
18 of 34
what is the calculated value?
number you get from the stats tet
19 of 34
to check for stat significance what do you have to do with this?
compare to critical value
20 of 34
what is the critical valy?
number that tells us whether or not we can reject the null and accept alternative
21 of 34
- lower levels of significance
dfgd
22 of 34
occasionally more stringent significance level used like 0.01 why?
may be human cost like drug trial or 'one-offs' thay couldnt be repeated
23 of 34
in all research if there's a large difference between calculated / critical what will happen?
researcher will check more stringent levels
24 of 34
and the lwoer the p value?
the more statistically significant
25 of 34
TYPE I AND TYPE II ERRORS
DFG
26 of 34
what is a type i error?
when null rejected and alternative accepted when should be other way round
27 of 34
what kind of error is this also referred to as?
optomistic error / false positive
28 of 34
why?
researcher claims to have fuond significant diff / correlation when one doesn't exist
29 of 34
what is a typeii error?
null accepted but should have been alternative bc that was true
30 of 34
also known as?
pessimistic error / false negative
31 of 34
more likely to make type i error when significance level is?
too lenient like 10%
32 of 34
and type ii more likely with significance?
too stringent like 1%
33 of 34
so why did we land @ 5% level?
best balances risk of type i / type ii error
34 of 34

## Other cards in this set

### Card 2

#### Front

- the null hypothesis

dfgdg

### Card 3

#### Front

what two things may hypotheses be?

### Card 4

#### Front

what may these two types of hypotheses also be referred to as?

why?