Summarising and presenting data Flashcards
What are the 2 main categories of data?
Qualitative / categorical
Quantitative / numerical
What types of qualitative data are there?
Unordered (male/female)
Ordered (small/ medium / large)
What types of quantitative data are there?
Discrete (number of children in family)
Continuous (SBP)
How would you examine the make-up of a sample of patients?
Using a frequency table
When is cross-tabulation useful?
Way of exploring whether two qualitative variables have any association (e.g sex and smoking)
What can bar charts and pie charts be used for?
To show proportions or percentages graphically
What is the median?
central value, such that 50% of the data falls at either side
What is the lower quartile?
Chosen to place 25% of the data below it
What is the upper quartile?
chosen to place 75% of the data below it
What is the inter-quartile range?
IQR
the difference between the 3rd and 1st quartiles, and quantifies the spread of the sample.
What is the mean?
1/n(sigma)xI
What is variance?
s2 = 1/(n-1) (sigma) (xi-mean)2
How do the mean and median relate to each other when the sample of the data have a reasonable symmetrical distribution?
mean = median
How do the mean and median relate to each other when the sample is positively skewed?
Mean >median
How do the median and mean relate to each other when the sample is negatively skewed?
Mean < median
What is a histogram a representation of?
The frequency distribution
If a histogram is shifted to the right it indicates?
Negative skew
If a histogram is shifted to the left it indicates?
positive skew
When is a doxplot useful?
When there are not many data points
How is positive skew seen in a box plot?
the distance from Q1 ro the median is much less from the median to Q3
What graph should be used to compare to quantitative variables directly?
scatterplot
When shouldn’t the mean be used to indicate the central value of the sample?
When the data is heavily skewed
When are box-plots inappropriate?
With very small sample sizes (3-4 points)
What are response variables?
these measures are used to describe the condition of an individual/subject (also called outcome measure or dependent variables?
What are explanatory variables?
Those measures that might explain the condition of an individual/subject (also called predictor or independent variables)
What are confounder variables?
Those measures that might obscure the relationship between response and explanatory variables. Associated with both the response variable and the explanatory variables
