Measures

Question 1

what is data ?

Answer 1

Data are measurements of any sort

Question 2

Types of data ?

Answer 2

Quantitative – number based

Qualitative - Words, concepts themes

Question 3

Types of quantitative data ?

Answer 3

Numerical

Categorical [ordinal or nominal]

Question 4

WHAT DO WE DO WITH DATA?

Answer 4

Data must be summarised and organised
Data can then be investigated and data from different groups
compared
 Identify themes and trends
 Describe variability
 Assess if differences exist
 Statistics – summary and manipulation of data

Question 5

what are quantitative measures ?

Answer 5

Numerical measures
 Should be done with a calibrated instrument or standard definition of measure
 Tape measure, amount of radiation in x-ray, w/cm2

Interval in nature
 Equal and logical intervals
 Continuous [can include parts of a whole, eg mA]
 Discrete [whole numbers, eg patient visits]

Question 6

what is frequency distribution ?

Answer 6

Frequency distribution aka histogram : graphical
representation of the number of times a value occurs
in a data set.

Question 7

what is normal distribution ?

Answer 7

Normal distribution will be bell shaped [Gaussian],

with data equally distributed from a central point

Question 8

what is abnormal distribution ?

Answer 8

Abnormal distribution has a skewed distribution of

data.

Question 9

Tests used for assessing distribution ?

Answer 9

Kolmogorov-Smirnov, ShapiroWilks

Question 10

when is Kolmogorov-Smirnov test used?

Answer 10

Kolmogorov -smirnov : for larger data set.

Question 11

When is ShapiroWilks test used?

Answer 11

Shapiro wilks- for smaller data set.

Question 12

For it to be normally distributed data, Significant test result should be ?

Answer 12

For it to be normally distributed the significant test result should be greater than 0.05

Question 13

What are the measurements of the central tendency ?

Answer 13

Mean value or average is the sum of all measures / number of measures.
If normally distributed data, this point represents from where the data
are half above and below.

Median is the value which occurs halfway when data is ranked in order [often used in abnormal distribution with quartiles and interquartile range]

Mode is the most commonly occurring value in the
data

Question 14

What is Standard deviation?

Answer 14

Standard deviation will describe the shape of the data set.

SD=average of the distance is from the central point.

Question 15

What is standard error ?

Answer 15

Standard error is used to approximate certain amount of data within the mean population.

SE=SD/√ n
Standard error is used to report the confidence limits.

95% cL= mean +/- 1.96 SE
95% Confidence limits : This explains where 95% of the data is sitting.
96% of the data within 2 SD of the mean.
95% is 1-1.96SE

Question 16

what is categorical data ?

Answer 16

Ordinal – ranked or ordered data, eg muscle strength, quality of life
Nominal – non ranked data, eg living arrangement, gender
Can treat this data categorically and measure percentages, and set
95%CL to approximate population distribution [assuming low type II
error or external validity]

Question 17

How is categorical data different from numerical data ?

Answer 17

categorical data :
It is subjective
The intervals are not equal and logical intervals occurring for the data