1/49
Looks like no tags are added yet.
Class :: A set, category, classification, or grouping; each row of a summary table is one class
Categorical data :: Labels or groups (e.g., payment type: cash, credit, debit, digital wallet)
Quantitative data :: Numbers you can do math on (e.g., average high temperature on 20 days)
Summary tools for categorical data :: Frequency table, relative frequency, bar chart, pie chart
Summary tools for quantitative data :: Class table, cumulative frequency, histogram, stem-and-leaf display
Summary tools for two variables :: Crosstabulation (two categorical variables) and scatter plot (two quantitative variables)
Frequency :: The count of items in a class; you get it by counting
Relative frequency :: Class frequency ÷ n; the share of the data in a class (0 to 1)
Percent frequency :: Relative frequency × 100; the same share as a percentage
Cumulative relative frequency :: Previous cumulative value + this class's relative frequency; the share of data at or below this class
Frequency table quick checks :: Frequencies add up to n, relative frequencies add up to 1.000, percent frequencies add up to 100%
Relative frequency of Cash (24,300 of 177,500 purchases) :: 24,300 ÷ 177,500 = 0.137, which is 13.7%
Bar chart :: Shows a frequency distribution of categories; classes along the bottom, frequency up the side; best for comparing counts between classes
Pie chart :: Shows a relative frequency distribution of categories; no axes, each slice is a share of the whole; best for showing percentages to the general public
Bar chart vs. pie chart :: Bar chart shows counts (frequency); pie chart shows shares (relative frequency)
Class table for quantitative data :: First group values into classes (ranges), then count and calculate shares; adds a cumulative relative frequency column
How do you build a quantitative class table? :: 1) Pick equal-width, non-overlapping classes that cover every value. 2) Count values per class. 3) Divide each frequency by n. 4) Add each relative frequency to the previous cumulative value; the last must equal 1.00
How do you read a cumulative relative frequency of 0.80 in the 80–89 class? :: 80% of the days had a high of 89°F or lower
Histogram :: The bar chart for quantitative data; graphs the frequency distribution and reveals the shape of the data
How is a histogram different from a bar chart? :: Histogram bars touch because the classes run continuously; a bar chart for categories has gaps
Symmetric histogram :: Peak in the middle, both sides mirror each other (bell shape)
Skewed left :: Peak on the right, long tail stretching left toward low values
Skewed right :: Peak on the left, long tail stretching right toward high values
Memory trick for skewness :: The skew is named for the direction of the tail, not the peak
Stem-and-leaf display :: Shows the shape of the data like a histogram but keeps every original value
Stem :: The leading digit(s) of a value in a stem-and-leaf display
Leaf :: The last digit of a value in a stem-and-leaf display
How do you build a stem-and-leaf display? :: 1) Sort data smallest to largest. 2) List stems down the left and draw a vertical line. 3) Write each value's last digit as a leaf next to its stem. 4) Keep each row's leaves in order
How do you read the row 8 | 2 5 5 7 9 9 9 9 9 ? :: It means the values 82, 85, 85, 87, 89, 89, 89, 89, 89
Stem-and-leaf display on its side :: Looks like a histogram; the longest row is the tallest bar
Crosstabulation :: A table that summarizes two categorical variables at once (e.g., payment type by company)
Cell relative frequency in a crosstab :: Cell count ÷ grand total; the share of all data in one crosstab cell
What is the grand total in the payment crosstab? :: 177,500, the same n as the payment-type table
Cash at BK relative frequency :: 10,300 ÷ 177,500 = 0.058, so 5.8% of all purchases were cash at BK
Biggest cell in the payment crosstab :: Credit at DB (37,200 purchases, or 0.210 of all purchases)
What do the Total row and column of a crosstab show? :: Each is the same kind of summary for one variable alone (right column = payment type, bottom row = company)
Two quantitative variables :: Plot each (x, y) pair as a point with one variable on each axis
Demand curve :: A graph where, as price goes down, the quantity people buy goes up
Connect the dots (predicting from a graph) :: Find the value on one axis, go across to the line, then drop to the other axis to predict a value that was never measured
Price vs. quantity line from the class example :: Price = 10 − 0.005 × Quantity; at a price of $7 the predicted quantity is 600
Why does connecting the dots matter? :: It lets a manager predict values that were never measured; exact for a perfect line, only an estimate with real data
Which summary table fits one categorical variable? :: Frequency and relative frequency (graph with bar chart or pie chart)
Which summary table fits one quantitative variable? :: Classes with frequency, relative frequency, and cumulative relative frequency (graph with histogram or stem-and-leaf)
Which summary fits two categorical variables? :: Crosstabulation
Which summary fits two quantitative variables? :: List of (x, y) pairs, graphed as points and connected to predict
54,000 of 177,500 purchases were debit. What is the percent frequency? :: 54,000 ÷ 177,500 = 0.304, so 30.4%
Which chart shows relative frequencies of categories to the public? :: A pie chart
A histogram peaks on the right with a long tail on the left. What is its shape? :: Skewed left
In the temperature table, what does 0.35 in the 70–79 row mean? :: 35% of days had a high of 79°F or lower
In the crosstab, what does 0.210 for Credit at DB mean? :: 21.0% of all purchases were credit card at DB