Variables, Levels of Measurement and Descriptive Statistics Flashcards
Operationalise
making it measurable
What level is Categorical
most basic level - also called nominal
categorical
assigning people or things to a category normally assigned a number
nominal numbers
No value - cant perform maths on them
binary variable
nominal categories with only 2 possible categories
data collected from categorical
frequency
ordinal
ranking items - size of number/differences between numbers have no meaning - no maths
interval
assume equidistant points between each of the scale
elements. eg IQ
interval maths
addition and subtraction
Ratio
like the interval scale in that there are equal
intervals between points on the scale.
* in addition to this the ratio scale has a true zero
Ratio maths
all mathematical operations
central tendency measures
mid point
mean
average - “arithmetic mean”
median
middle value
mode
most common value
levels of measurement mean can be applied to
interval and ratio only appropriate for continuous data
levels of measurement median can be applied to
ordinal,interval and ratio
levels of measurement mode is appropriate for
all levels
effect of outliers: mean
very sensitive
effect of outliers: median
insensitive
effect of outliers: mode
insensitive
outliers
scores that are numerically very far apart from the rest of the data you have
in a particular set
range
difference between lowest and highest scores in a data set
-ves of range
- only describes extreme scores
- sensitive to outliers
- tends to increase with sample size
interquartile range
only looks at middle 50% of scores
+ves of IQR
- It gives the central grouping of the values in a data set.
* It is the best measure of dispersion for ordinal data.
-ves of IQR
• Does not use all the scores in a set in its calculation
variance
based on the difference of each score from the mean of those scores
+ves variance
based on every score
-ves variance
obtained value is not immediately meaningful because it is not in the original units of the data (need to work out square root)
standard deviation
the average deviation between a set of scores and their mean - bigger the SD = bigger spread of data
+ves of SD
- based on all scores
- obtained value is meaningful in terms of the original units of the data
- best measure for interval/ratio data
-ves of SD
- not meaningful for ordinal
* sensitive to extreme scores
