2.1 Representations of Data - Topic 2 Data Presentation and Interpretation Flashcards
Qualitative variables/data:
non-numerical data that come in classes or categories e.g. favourite colours, makes of car
Quantitative variables/data:
numerical data for which the numbers are meaningful e.g. tines to run a race, height
Discrete data:
set of values of data can be listed e.g. shoe size, number of goals, number of children in family etc.
Continuous data:
values can’t be listed because data can take any value in a particular range e.g. height, mass, time etc.
What can a histogram represent?
- groups continuous data can be presented using histograms
- histograms show rough location, general shape of data and how spread out data is
Formulae used for histograms:
area of bar = k x frequency
- k=1 is easiest value to use when drawing histogram
- if k=1 then frequency = freq. density x class width
area (k x frequency) = freq. density x class width
2 ways of calculating frequency density for histograms:
frequency density = frequency / class width
frequency density = frequency / number of standard widths
- if class widths have common factor then it is often convenient to take this common factor to be the standard width
- vertical axis must be labelled accordingly e.g. frequency per ‘kg’
How do you use number of standard widths in histograms to calculate frequency?
- if common denominator of class widths is 5 calculate f.d. = frequency/5
- calculate frequency per 5mins by diving frequency by number of standard units
- e.g. if class width is 10 the number of standard units would be so to find frequency per 5 minutes = frequency / 2
- area of a bar would now be area = 5 x frequency
Histogram example:
What are box/whisker plots used to show?
- quartiles, max and min values and any outliers
- box drawn from LQ to UQ = 50% of data
Box/whisker plot diagram:
What does the area in a bar in a histogram represent?
frequency
What does cumulative frequency enable us to do?
- enables us to estimate how many items of data fell below any particular value
- used to estimate medians, quartiles and percentiles of the data
How must cumulative frequencies for grouped data be plotted?
against the upper class boundaries
How do you calculate the median, UQ and LQ from cumulative frequency graph?
- if frequency is 100
- median is 100/2 = 50th value -> look on graph what x value the 50th value has
- same for LQ and UQ
Rules for drawing frequency polygons:
- calculate midpoint of each class width
- plot class frequency at midpoint for the class
- connect the plotted points using straight lines
Frequency polygon diagram:
Frequency polygon on a histogram diagram:
How do we compare 2 data sets
Comment on:
- measure of location
- measure of spread