Research methods C Flashcards
what is the p in statistics
the likelihood that we are wrong about nothing going on
example of when unethical research has been published
august 2022, researcher immersed himself in porn featuring children
what is a rationale
an explanation of the reasons of doing something
conceptual (in terms of a hypothesis)
describing an idea clearly and concisely using accessible language
technical (in terms of a hypothesis)
describing an idea precisely and with reference to appropriate technology
what is the structure of a simple comparison experiment
1 IV, 1 DV, 2 conditions
what is the structure of a one-way design
1 IV, 1 DV and 3 or more conditions
how many differences would you see in a one-way design?
there are 3 differences in a one-way design
how is a factorial design different?
looking at more than one IV at the same time
does the number of levels have an effect on interactions?
no the number of levels do not have an effect on interactions
what do the number of factors determine? (in a factorial design)
the number of interactions
what does ANOVA stand for
analysis of variance
what is variance
measure of spread/dispersion
what does ANOVA test
are two or more groups from the same population of scores - are the differences in individual scores the same size?
why is there variation in scores?
naturally occurring differences (error) ,differences due to a factor that varies for different groups (known as main effects) and the interactions
what is the definition of error?
the naturally occurring differences between people
where is the mean on a bell curve
at the peak
what does this show
same mean, different levels of dispersion
what does the width of a singular bell curve show
within-group variance
what does the length between the peak of two bell curves mean
the between group variance
what does it mean if the f ratio is bigger than 1
significant difference
what is the f ratio
the ratio of between group variance and within group variance
between group variance
the overall variation between the groups
within group variance
the overall variation in one experimental group
when is a post hoc test used
when an ANOVA f test is significant
type 1 error
concludes there is a difference when there isn’t
what type of data is used for ANOVA
normally distributed interval and ratio data
if there are two groups what are the chances of making a type 1 error
1/20
how do you get around the family wise error
set an acceptable error rate
what do you report for a non-parametric test used by ANOVA
medians and minimum&maximum scores
what is a kruskal-wallis test
a non-parametric independent groups one-way ANOVA
how many groups does a kruskal-wallis test compare
3 or more independent groups
what are the two non-parametric tests used to test the difference between three or more conditions?
kruskal-wallis and friedman
what is a Friedman test
a one-way non-parametric repeated measures ANOVA
if parametric assumptions were violated by ANOVA what do you do?
use a non-parametric test - kruskal-wallis or Friedman
what does the ‘number of ways’ influence (in relation to ANOVA)
the number of IV’s
how many IV’s does a three way ANOVA have
3
do the number of ‘ways’ in an ANOVA influence the number of DV’s
no
how do you work out the number of conditions in an ANOVA
number of factors x number of levels
how many conditions if there are 2 factors and 3 levels
6
how many interactions for a 2 x 5 ANOVA
1 interaction as there are only two conditions
why do we need to carry out post-hoc tests
because a signficant ANOVA does not show what conditions are different to one another
what is the bonferroni correction
takes the number of comparisons and takes the acceptable p value, divides it by the number of tests that’s going to be used
formula for working out number of interactions in an ANOVA
how do you know in a graph whether two factors have an interaction?
lines should be parallel
