1What does descriptive analytics answer?
What happened, by summarizing past data
2What does predictive analytics answer?
What will happen, by estimating an outcome
3What does prescriptive analytics answer?
What we should do, by selecting an action to meet an objective
4“What was last quarter’s average sales?” is which type of analytics?
- Descriptive
- Predictive
- Prescriptive
5“How many units will we sell at \$5.49?” is which type of analytics?
- Predictive
- Descriptive
- Prescriptive
6“What price maximizes revenue?” is which type of analytics?
- Prescriptive
- Descriptive
- Predictive
7What does validity ask?
Whether a measure represents the concept it is meant to measure
8How does validity differ from reliability?
Reliability is accurate, consistent recording; validity is whether the right concept is measured
9Why is a perfectly counted number of customer service calls an invalid measure of dissatisfaction?
Many calls are just questions, so the count doesn’t measure dissatisfaction
10What is a model?
A simplified representation of a real situation
11What does a model keep and leave out?
It keeps the important features and leaves out the rest
12What are the three forms of a model?
Verbal, visual, and mathematical
13What is a verbal model?
A description in words
14What is a visual model?
A graph or sketch
15What is a mathematical model?
An equation or spreadsheet
16What are the inputs of a simple cost model?
Fixed cost, variable cost per unit, and quantity
17What is the result of a simple cost model?
Total cost
18What is the total cost formula?
\$\text{Total cost} = \text{Fixed cost} + (\text{Variable cost per unit} \times \text{Quantity})\$
19What is total cost with fixed cost \$50,000, variable cost \$125 per unit, and quantity 1,500?
\$237,500
20What does a \$ in an Excel cell reference do?
It locks whatever follows it when the formula is copied
21What happens to a part of a reference with no \$ when copied?
It shifts by the same number of rows and columns moved
22How does a relative reference such as A1 behave when copied?
Both column and row shift
23How does an absolute reference such as \$A\$1 behave when copied?
Nothing shifts
24How does a mixed reference \$A1 behave when copied?
Column is fixed, row shifts
25How does a mixed reference A\$1 behave when copied?
Row is fixed, column shifts
26What does =A2+B\$1 in C2 become when copied to D3?
=B3+C\$1
27In a database table, what are records?
Rows
28In a database table, what are fields (attributes)?
Columns
29What does a filter do to records that don’t match?
It hides them; it doesn’t delete them
30Why can a sum over filtered data be wrong?
Highlighting the range still includes the hidden rows
31How do you compute on only the visible filtered rows?
Copy and paste the visible rows elsewhere
32Which filter shows orders with item cost of at least \$200?
Numbers Filter, Greater Than Or Equal To, 200
33What is the syntax of the IF function?
\$\text{IF}(\text{condition}, \text{valueiftrue}, \text{valueiffalse})\$
34When does AND return TRUE?
Only if every test is true
35When does AND return FALSE?
If any one test is false
36What does =IF(AND(B2>=10000,C2>=25000),”Critical”,”OK”) return with B2 = 12,000 and C2 = 20,000?
”OK”, because the second test is false
37When do you use VLOOKUP?
When the table is arranged vertically, with lookup values down the leftmost column
38When do you use HLOOKUP?
When the table is arranged horizontally, with lookup values across the top row
39What are the four arguments of VLOOKUP?
Lookup value, table range, return column index, and exact-match setting (FALSE)
40How is the VLOOKUP column index counted?
From the left edge of the selected range, not the worksheet column letter
41In range B2:E8, which column does a VLOOKUP index of 3 return?
Column D
42What does FALSE do in VLOOKUP?
Requires an exact match
43What does VLOOKUP return if no exact match is found with FALSE?
#N/A
44What does omitting the last VLOOKUP argument, or using TRUE, require?
A first column sorted in ascending order
45Why use an absolute table range like \$A\$2:\$E\$8 in VLOOKUP?
It keeps the range locked when the formula is copied down
46What does Data Validation do?
Restricts what can be entered in a cell
47What are two examples of Data Validation restrictions?
A list of permitted categories, or whole numbers from 1 to 4
48Why does Data Validation help data quality?
It prevents typos and inconsistent entries that break filters, COUNTIFs, and PivotTables
49Which PivotTable areas do you drag fields to?
Rows, Columns, and Values
50For revenue by region and product, which field goes in Values?
Amount
51What does Sum do in a PivotTable?
Adds the numbers
52What does Count do in a PivotTable?
Tallies how many records
53Which PivotTable aggregation answers “How many transactions?”
54Which PivotTable aggregation answers “Total revenue?”
55What is the default PivotTable value summary?
Sum
56Where do you change a PivotTable value summary?
Value Field Settings
57Which chart shows change over time?
Line chart
58Which chart compares values across categories?
Bar or column chart
59Which chart shows the relationship between two numerical variables?
Scatter chart
60How do you justify a chart choice?
Name the data type and the business question the chart answers
61What does a truncated chart baseline do?
Exaggerates differences between values
62Why start a bar chart axis at zero?
It makes bar heights proportional to the values
63What are data bars?
Conditional formatting that puts a colored bar inside each cell, scaled to its value
64What does a longer data bar mean?
A larger value
65What is a dashboard?
A visual set of key business measures that lets managers monitor performance and spot problem areas
66What is the mean formula?
\$\text{Mean} = \dfrac{\text{Sum of the values}}{\text{Number of values}}\$
67What are the mean and median of 10, 12, 13, 15, 100?
Mean 30, median 13
68How does an extreme value affect the mean versus the median?
It pulls the mean toward it but barely affects the median
69Which measure of a typical value is better when data contain an extreme observation?
The median
70What is the kth percentile?
A cutoff with about k% of observations at or below it
71Which percentiles are Q1 and Q3?
Q1 is the 25th percentile and Q3 is the 75th percentile
72What is the IQR formula?
\$\text{IQR} = Q_3 - Q_1\$
73What is the IQR if Q1 = 40 and Q3 = 65?
25
74What does the IQR measure?
The spread of the middle half of the data
75What does a smaller standard deviation mean?
Less spread, when compared in the same units
76Can two groups have the same mean but different variability?
Yes, for example mean 50 with s = 2 versus s = 10
77Which is riskier, Intel (s = \$0.50) or GE (s = \$0.35)?
Intel
78What is the z-score formula?
\$z = \dfrac{\text{value} - \text{mean}}{\text{standard deviation}}\$
79What is the z-score for value 85, mean 70, s = 5?
+3.0
80What is the z-score for value 62, mean 70, s = 4?
-2.0
81What does the sign of a z-score tell you?
Whether the value is above (+) or below (−) the mean
82What does the magnitude of a z-score tell you?
How many standard deviations the value is from the mean
83What should you do first with an unusual observation?
Check the original record and business context
84How should a data error such as a typo be handled?
Correct it
85How should a legitimate unusual value be handled?
Keep it, but handle it with care in the analysis
86Should you delete a value just because it is extreme?
No
87What is the home value regression equation from the slides?
\$\text{Market value} = 32{,}673 + 35.036 \times \text{square feet}\$
88In the home value regression, what is x (the predictor)?
Square feet
89In the home value regression, what is the predicted outcome?
Market value in dollars
90How do you interpret the slope of 35.036?
Each additional square foot is associated with about \$35.04 higher predicted market value
91What is the regression prediction formula?
\$\text{Predicted outcome} = \text{Intercept} + (\text{Slope} \times x)\$
92What is the predicted value for 1,750 sq ft?
About \$93,986
93What is the predicted value for 2,200 sq ft?
About \$109,752
94Why isn’t the intercept meaningful in the home value regression?
x = 0 is outside the data range of about 1,400-2,600 sq ft
95Does regression establish cause and effect?
No, it shows association only
96What is R Square in the home value regression?
0.5347, about 53%
97How do you interpret R Square of 0.5347?
53% of the variation in market value is explained by home size, in this sample
98Is R Square the percent of predictions that are correct?
No
99Does R Square show the direction of the relationship?
No
100What is the residual formula?
\$\text{Residual} = \text{Actual value} - \text{Predicted value}\$
101What is the residual for actual \$90,000 and predicted \$96,159.13?
-\$6,159.13
102What does a positive residual mean?
Underprediction (actual is higher than predicted)
103What does a negative residual mean?
Overprediction (the model predicted too much)
104In what units should residual error be reported?
The outcome’s units (dollars in the home value example)