Statistics 1

Question 1

Why do we screen data?

Answer 1

1. to detect blunders (value which is obviously incorrect)
2. to locate outliers (plausible but could influence results a lot)
3. distributional properties
4. missing values

Question 2

How do we screen data?

Answer 2

Small data set - eye ball the data
Large data set - frequency table or histogram

Question 3

What do we do about blunders and outliers?

Answer 3

Blunders - correct
Outliers – take into account e.g. non-parametric

Question 4

What are the different scales of measurement?

Answer 4

Qualitative (categorical)
- nominal (no order) e.g. male/female
-ordinal (is an order) e.g. pain- mild, moderate, severe
Quantitative
-interval (scale) e.g. shoe size

Question 5

How are categorical values presented? (qualitative)

Answer 5

Tables of frequencies/relative frequencies
Graphically in bar charts (there are gaps between the bars)

Question 6

How are interval scale values presented? (quantitative)

Answer 6

Histograms (no gaps between bars)

Question 7

What is the variance?

Answer 7

variance is the standard deviation squared

Question 8

Describe a boxplot

Answer 8

Boxplot- good for comparing groups
Shows min and max value, quartiles and median

Question 9

Discuss the mean

Answer 9

Coincides with the median if the distribution of the data is symmetrical.

Question 10

Discuss the median

Answer 10

Is a useful summary measure when the data are skewed.
Median is equal to the 50th percentile

Question 11

What is dispersion?

Answer 11

How much scatter/spread there is in the distribution