Module 1

Question 1

What is an element?

Answer 1

an entity on which data is collected.

Question 2

What is the nominal scale of measurement?

Answer 2

labels, names, numerical codes - things we can’t math.

Question 3

What is the ordinal scale of measurement for?

Answer 3

Like nominal but there’s a ranking system (i.e.: good, better, best).

Question 4

What is the interval scale of measurement?

Answer 4

• Like ordinal but intervals are a fixed unit of measure.
• Always numerical.
• Generally have one numerical value in their calculation that doesn’t actually hold a lot of meaning (might take a categorical value into account).
• i.e. SAT scores

Question 5

What is a Ratio scale of measurement?

Answer 5

Like interval, but the ratio of two values is meaningful. I.e. distance, height, weight, time. Requires a 0 value.

Question 6

What are the two classes of data?

Answer 6

Categorical
Quantitative

Question 7

What scales are used for categorical data?

Answer 7

Nominal or ordinal

Question 8

What scales are used for Quantitative data?

Answer 8

Interval and Ratio.

Question 9

What are the two types of data series?

Answer 9

Cross-sectional
Time series

Question 10

What is time-series data?

Answer 10

Collected across time

Question 11

What is cross-sectional data?

Answer 11

collected at roughly the same point in time.

Question 12

What’s an observation?

Answer 12

The set of measurements obtained for each element is one observation.

Question 13

What’s the difference between a census and a sample survey?

Answer 13

Census = population
Sample survey = sample

Question 14

What are the three types of analytics?

Answer 14

Descriptive
Predictive
Prescriptive

Question 15

What are the three v’s of big data?

Answer 15

Volume
Velocity
Variety

Question 16

What’s a frequency distribution?

Answer 16

Tabular summary of how often each thing in each category happened.

Question 17

What’s the difference between a frequency distribution, a relative frequency distribution, and a percent frequency distribution?

Answer 17

Frequency = #
Relative frequency = proportion
Percent = %

Question 18

How do we create classes for data?

Answer 18

• Non-overlapping
• 5 - 20
• same width
• round the width of the class so it’s not a ton of decimal points
• class limits: each data item should only belong in one class
• same number of digits as data
• use open-ended classes (either no upper or no lower limit) for the ends as a catch-all to simplify things

Question 19

How do we calculate the widths for a pie chart?

Answer 19

Relative frequency * 360 degrees

Question 20

What does a dot-plot look like?

Answer 20

Question 21

What does a histogram look like?

Answer 21

Question 22

What does a cumulative frequency distribution look like?

Answer 22

Shows the number of data items with values less than or equal to the upper class limit of each class, so each successive class includes all the ones before it.
Last number will always be the total.

Question 23

What does a stem-and-leaf display look like?

Answer 23

The first digit of the data point goes on the left, and the second digit for any data point that starts with the same first digit is put beside it. For bigger numbers use a “leaf unit” = 10, etc.

Question 24

What is a cross-tabulation?

Answer 24

• tabular summary for two variables.
  
  • May be different types of data.
  • Finds relationship between two variables.
    ***don’t forget the column and row totals!

Question 25

What is Simpson’s Paradox?

Answer 25

Data in two or more crosstabulations that are summarized may actually show a reverse relationship to what was shown when the data wasn’t aggregated.

Question 26

What is a side-by-side bar chart?

Answer 26

Question 27

What is a stacked bar chart?

Answer 27

Question 28

How do we create a display?

Answer 28

• Title
• Simple
• Axis labels
• Unit of measure
• Distinct colours
• Legend

Question 29

Which types of displays show distribution?

Answer 29

• Bar chart
• Pie chart
• Dot plot
• Histogram
• Stem-and-leaf

Question 30

What types of displays are used to make comparisons?

Answer 30

• Side-by-side bar chart
• Stacked bar chart

Question 31

What types of displays are used to show relationships?

Answer 31

• Scatter diagram
• Trendline

Question 32

What do we use to denote the mean for a sample?

Answer 32

xbar

Question 33

What do we use to denote the mean for a population?

Answer 33

mu

Question 34

Why is median sometimes preferred to mean?

Answer 34

Median is not influenced by outliers.

Question 35

When can we not use the mean?

Answer 35

Financial growth data - need to use the geometric mean for that.
When observations have different levels of importance - use weighted mean for that.

Question 36

What do s and sigma mean?

Answer 36

s = sample standard deviation
sigma = population standard deviation

Question 37

What are the traits of a symmetric distribution?

Answer 37

mean = median
skewness = 0

Question 38

What traits does a right - skewed distribution have and what does it look like?

Answer 38

mean > median
skewness > 0

Question 39

What traits does a left-skewed distribution have and what does it look like?

Answer 39

Mean < median
Skewness < 0

Question 40

What is a Standardized Value?

Answer 40

• z-score
• the number of standard deviations xi is away from the mean.

Question 41

What does a negative, positive and 0 z-score mean?

Answer 41

a) - = xi < mean
b) + = xi > mean
c) 0 = xi = mean

Question 42

How do we know something is an outlier?

Answer 42

z >3
z< -3
outside of Q1 - 1.5 IQR or Q3 + 1.5 IQR

Question 43

What’s in a five number summary?

Answer 43

Smallest
Q1
Q2
Q3
Largest

Question 44

What does a box-plot look like?

Answer 44