what are the inferntial statistal tests for nominal data
unrelated data: chi square test
related data: binominal sign test
what are the inferntial statistal tests for ordinal data
unrelated data: mann whitney test
related data: wilcoxon signed ranks test
correlation test: spear mans rho correlation coefficient
what are the inferntial statistal tests for interval or ratio data
unrelated data: independent t-test
related data: related t-test
correlation test: persons product moment
what are inferential statistics
they enable you to work out which hypothesis has been supported by the data from either an experiment or a correlation study
whats the three criterea that have to be met for using a parametric inferential statistic test
what are type 1 errors
alternitive hypothesis is accepted, null hypothesis is rejected and any behaviour shown is due to chance
what are type 2 errors
rejection of alternitive hypothesis and acceptance of null hypothesis when the IV has an impact on the DV
whats normal distribution
data about any behaviour from a representative sample of the population
a ‘true’ normal curve is when all measures of central tendancy occur at the highest point in the curve
mean=mode
If your data is normally distributed, parametric tests are…
parametric tests are more powerful (better at detecting real effects).
If your data isnt normally distributed, parametric tests can…
give misleading results, so you must use a non-parametric alternative.
whats negatively skewed distribution
the mean is different from the mode and there is more or fewer low scores or high scores
mean<mode
whats postievtly skewed data
if the population has more or low scorers than highers
mean>mode
what tests are parametric
independent t-test
related t-test
persons product moment
whats tests are non parametric
chi square, mann whittney, wilcoxon signed ranks, binominal sign, spear mans rho
if the standard of significance isnt reached how do psychologists express this
p>0.05
null hypothesis supported
if the standard of significance is reached how do psychologists express this
p<_0.05
alternitve hypothesis is supported
if the standard of significance is exceeded how do psychologists express this
p<_0.01
alterntive hypothesis strongly supported
whats the split half method
tests internal reliabilty
tests one half of the questions and gains a score then the other half of Q’s
see if same level of score was acheved on both halves
whats test retest method
tests external reliability
high level of standardisation would make research replicable
same results on a retest will show external relaibilty
inter rater relaibility
avoids observer bias
two observers consistently rate or observe the same behaviour and the two sets of ratings are correlated
if a positive correlation is seen, inter rater reliability has been established
whats the harvard method
Name
Date
Title
J
Volume
Issue
P
refereces that include 7 things
the author
date of publication
issue
volume
title
publisher
whats the abstract in a report
summary of the research outlining the aim, method, results etc
its in the first section of the report but written at the end
whats the method in a report
lenghty section which includes all details for someone to carry out the same research
DETAILS OF SAMPLE OBTAINED
whats a discussion in a report
focuses on what the study has discovered
evaluations, alternitve explanations and references to locate any info included