qualitative revision Flashcards
independent varibale
the one we manipulate
dependent variable
the one we hypothesise about what we measure
nominal data
variables with different values.
(eg: green eyes)
yes and no
ordinal data
data can be ordered in a scale (1st/2nd/3rd)
interval data
set in stone values (IQ)
sampling
portion of the population
population
a group of persons that have a common characteristic
what sampling method do you use to represent a population
random sampling
in reality what samplin is done
take whoever volunteers
to test a hypothesis what 2 things do we need to compare
pre v post group
treatment v control group
independent mesures
two sperate group
paired meaure
one group measured twice
central tendency
mean
median
mode
measires of varience
standard deviation
histograms
standard error
normal distribution
positive up the ladder to slide
negative walk up the slide
probability
0-1 scale
0-100%
p=0.50
what does p = 0.05
5% chance of something happening
what do journal articles report at
less than 5 p<0.05
what do we use inferentual statistics for
-objectively interpret data
-make inferences about popualtion
-draw conclusions
what is a null hypothesis
no difference in the population
any numerical difference is by chance
what is the highest p value we reject
alpha - 0.05
examples of ratio variables
weight/height/vo2max
what is validity
the study measures what you actually tend to measure
how do we measure varience
standard deviation
what is statistical power
probability of something being correct
nonparametric tests
wilcoxon and mann-whitney U test
can be used with any distribution
-ordinal
-interval
-ratio
parametic tests
use even more information meaning they have greater power
sign test parametric what more info does it use
are scores higher or lower
-+
signed rank test (wilcoxon)uses
-are scores higher or lower
-are the differences greater for the “higher” scores for the “lower” scores
t test
-are scores higher or lower
-more higher or more lower then get the difference
-assume that data is normally distributed
requirements for parametric test
-data has to be normally distributed
-data habe to be inteval or ratio
-the varience of the sample must be similar
2 types of varibality
-between subject variability
-witin-subject variability
between subject variability
-different subjects will have different scores for a particular variable
-there is variablity BETWEEN subjects
within-subject variability
-when you measure a given subject more than once, you will get different results.
-but this variability is usually less than between subjects
