Inferential Testing Flashcards
what are inferential statistics?
when you use data collected from sample to make assumptions (infrences) about behaviour of entire target population
when is a result deemed significant?
probability
if the probability is less than 5% (or 0.05).
what happens if the result is not deemed significant?
probability
the null hypothesis is accepted and alternate is rejected.
when do you use the sign test?
the sign test
- when a test of difference
- when repeated measures or mathced pairs used (related data)
- when data is nominal
how do you conduct the sign test?
the sign test
- subtract each ppts score from condition 1 with score from condition 2.
- add a + or a - sign depending on if scores went up or down.
- work out N (add up the number of ppts where there was a difference in the conditions)
- work out S (the least frequent sign)
- use a critical values table, reading across from N
- is calculated S value is less than or equal to the critical value it is significant
- OR find the N and the S and look for the percentage probability. if less than 5% significant
what is H0?
probability and significance
the null hypothesis - suggests no causal relationship between IV and DV
what is H1?
probability and significance
the alternate hypothesis - suggests causal relationship
what must a researcher decide after collecting data?
probability and significance
is it strong enough to accept the hypothesis? natural variablity will always occur even in highly controlled studies.
what can researchers use to minimise chance of choosing wrong hypothesis?
probability and significance
probability
what is the agreed probability that result is due to chance?
probability and significance
greater than 0.05 (aka 5% or 1/20)
if less than 5%, behaviour is not due to chance and is significant
sometimes 0.01 is used for controversial theories.
what is a type I error?
probability and significance
when researchers accept the alternate hypothesis and reject the null hypothesis in error. a false positive
what is a type ii error?
probability and significance
when researchers accept the null hypothesis and reject the alternate hypothesis in error. chance hid the causal R so appeared to not be significant
how do you reduce type i errors?
probability and significance
use P=<0.01
this will increase likelihood of type ii though.
how do you reduce type ii errors?
probability and significance
use P=<0.05
will increase likelihood of type i though.
what are the statistical tests used in psychology?
inferential tests
Chi Squared
Sign Test
Mann Whitney
Wilcoxon
Spearman’s Rho
Unrelated T Test
Related T Test
Pearson’s
