WEEK 3 - SUMMARISING DATA

What are descriptive statistics?

describe and summarise data

What are inferential statistics?

results of statistical tests, allow us to draw conclusions from the data

What is absolute frequency?

amount of each value/response

What is relative frequency?

proportion of each value/response

What graph is used to visualise frequencies?

histogram

What type of histograms plot absolute values?

frequency histograms

What type of histograms plot relative frequency values?

relative histograms

What type of histograms plot cumulative frequency values?

cumulative histograms

What is a normal distribution?

scores are distributed around a mid-point (mean)

What is a skewed distribution?

there is a bulk of responses at either end of the scale

What is a positive skew?

value cannot be below 0, so a majority of responses are on the low end (LEFT) of the scale

What is a negative skew?

value cannot be above 1, so a majority of responses are on the high end of the scale (RIGHT)

What are floor and ceiling effects?

when a task is too difficult or simple to capture variation in ability

What is central tendency?

a single value that represents the whole distribution

What measures of central tendency are there?

mode, mean, median

What is dispersion?

reflects how far from the mean the average data point is

How is dispersion calculated?

sum of squared differences, variance, standard deviation

How is sum of squared differences calculated?

calculate each data point’s distance from the mean, square them, and sum them

How is variance calculated?

sum of squares / (number of observations - 1)

How is standard deviation calculated?

square root of variance

Why standardise scores?

to see where a particular score sits within the overall sample

What is a Z-score?

reflects where a given score lies within the distribution (how many SDs from the mean)

How is z-score calculated?

Z = (score - mean) / (standard deviation)

What proportion of data points can be expected to lie within 1 standard deviation of the mean if the data is normally distributed?

68.2%

What proportion of data points can be expected to lie within 2 standard deviations of the mean if scores are normally distributed?

95.4%