Epi 3 - Descriptive Statistics

Question 1

Descriptive Statistics

Answer 1

Non-comparative, simple descriptions of various elements of study’s data

Provides readers w/ description of data, study subjects (patients), and elements about them

Question 2

What do the Measures of Central Tendency describe about data set?

Answer 2

Describe the dispersion or spread and variance + consistency (variance + SD) of all data

Mean, Median, Mode
Minimum, Maximum, range
Interquartile Range
Variance
Standard deviation

Question 3

Which Measures of Central Tendency describe the consistency and variance of data?

Answer 3

variance

standard deviation

Question 4

Which of the REQUIRED ASSUMPTIONS of interval/ratio data for proper use in Parametric Test can be determine by descriptive statistics?

Answer 4

Normally-distributed

Normal distribution of data can be determined 3 different ways via descriptive statistics

Question 5

Parametric Test =

Answer 5

All statistical data tests for Interval/Ratio data

Question 6

3 ways to check for symmetry or normal distribution in data

Answer 6

1) Mean/Median
2) Look at shape of Graphical Representation
3) Skewness value or Kurtosis value; value 0 = normal distribution both

Question 7

Mean

Answer 7

Average all data

Only significant in interval data

Question 8

Median

Answer 8

The middle value of data set when in ascending order

Question 9

Mode

Answer 9

Highest frequency value of data set

Question 10

Interquartile Range

Answer 10

Middle 50% of data values of normally-distributed data

Between 25th - 75th percentiles

Question 11

If normal-distribution, the mean/median of data =

Answer 11

~ 1

Mean & median = equal or near equal values

Question 12

Data distribution is skewed anytime

Answer 12

Median DIFFERS from Mean

Question 13

mean > median =

Answer 13

Positively Skewed Distribution Data - skew to RIGHT

• Mean LARGER than Median
• Asymmetrical curve = longer ‘tail’ extends/points to right of graph (extends positive values graph)
• Skewness and Kurtosis NOT = 0

Question 14

mean < median =

Answer 14

Negatively Skewed Distribution Data - skew to LEFT

• Mean SMALLER than Median
• Asymmetrical curve = longer ‘tail’ extends/points to LEFT of graph (toward negative graph values)
• Skewness / Kurtosis NOT = 0

Question 15

mean = median

Answer 15

Normally-distributed data

Symmetrical bell curve with data even dispersed either side of mean

Question 16

Skewness

Answer 16

Statistical value that discerns whether or not data equally distributed.

Skewness value = 0 = normally-distributed data

Question 17

Kurtosis

Answer 17

Measure extent that the data clusters around mean value. Kurtosis statistic = 0 then data Normally-distributed

Fxn of SD: LOWER stand. dev. = MORE data clusters around mean; HIGHER stand. dev. = LESS data clusters around mean - more dispersed

Kurtosis value = 0 = normal distribution
+ value = more cluster around mean
- value = less cluster

Question 18

Describe relationship between standard deviation & kurtosis

Answer 18

Small standard deviation = more clustering = + kurtosis value
Large standard deviation = less clustering = - kurtosis

Question 19

Which descriptive statistic is used to describe and compare Continuous (Interval/Ratio data)? Why?

Answer 19

MEAN

B/C: Discrete data (Nominal/Ordinal) do NOT have consistency their scales of measurement / categories - thus - if mean = value not distinctly fall within their scale the value is irrelevant because it has no meaning
- ex.: Nominal data - gender (male = 0, female = 1) and mean = 0.6 – value has no meaning or relevance to scale
Ordinal data - questionnaire responses (never = 0, almost never = 2, sometimes = 3 almost always= 4, always = 5) and mean = 3.6 – value again is irrelevant because values between scale no meaning

Question 20

What descriptive statistic is useful for describing and comparing Ordinal Data?

Answer 20

MEDIAN

Provides a sense of the central, middle value data that is still relevant Ordinal data set because unlike mean, median value will fall within the scale of measure.

Median - central value when data in ascending order

Question 21

What percentile of data falls within 1 standard deviation of mean value of a normally distributed data set?

Answer 21

68%

+/- 34% either side mean value

Question 22

What percentile of data falls within 2 standard deviations of mean value of a normally distributed data set?

Answer 22

95%

+/- 13.5% either side 1 sd;
13.5 + 34 = 47.5% on either side of mean

Question 23

What percentile of data falls within 3 standard deviations of mean value of a normally distributed data set?

Answer 23

99.7%

+/- 2.35% either side 2 sd

Question 24

Variance and Standard Deviation descriptive values provide what information about data set?

Answer 24

Variability and Consistency