Descriptive Statistics Flashcards
What are descriptive statistics?
they described the data shown as a number which summarises the observed data
only tells us about sample
What are the 2 types of descriptive statistics
measure of central tendency
measures of dispersion (or spread)
What are the measures of central tendency?
mean, median, mode
what is a strength of the mode?
easy to calculate and not distorted by extreme values
what is a weakness of the mode?
focuses only on one value
what is a strength of the median?
less affected by outliers
what is a weakness of the median?
only uses 1-2 data points so doesn’t take into account all the data
what is a strength of the mean?
good representation of the data set due to taking all values into account
what is a weakness of the mean?
affected by extreme values, not useful for nominal data
What data level is linked to the mode?
nominal data
what data level is linked to the median?
ordinal
what data level is linked to the mean?
interval
what are the measures of dispersion?
range, interquartile range, variance, standard deviation
what is the range?
differences between the highest and lowest variable
what is the IQR?
you remove the upper an lower 25% of data, then complete the range in the middle 50% of scores
What is variance?
measure of how much values in a data set differ from the mean
How do you calculate variance?
find mean of data
find difference between each value and the mean
square each difference
add up squared differences
divide sum of squared differences by no. of data points
what is S.D?
square root of variance
Strengths and Weaknesses of range
+ easy to calculate
- effected by extreme values
- doesn’t take into account all data values
strengths and weaknesses of IQR
+ less effected by extreme values
+ good for ordinal data
- doesn’t use all the data values
strengths and weaknesses of S.D
+ scale invariance
+ uses all the scores
- time consuming to conduct
strengths and weaknesses of variance
+ treats all deviations the same
- sample size
normal distribution
symmetric, few extreme scores on each end, bell shaped curve
mean, median and mode close together
Skewed distributions
majority of scores on one end
extreme scores affect mean
Positively skewed distribution
data extends to right side
most data clustered on left
mode, median then mean extends off
Negatively skewed distribution
data extends to the left side of the most data clustered on the right
mean, median, then mode (right side)
What descriptive statistic would you use for interval/ratio data with a skewed distribution?
median
range
what descriptive statistic would you use for interval/ratio data with a normal distribution?
mean
standard deviation
