Unit 1, Ch 1 Flashcards
variables n shit
Frequency
num of times a variable occurs
Ex. 3 people like math, 3 people like English
Proportion
percent of frequency out of 1
Ex. 0.5 like math, 0.5 like English
2 traits to watch out for in a chart (y, pic)
- how the y axis is labeled
- using pictures instead of bars
Things to ALWAYS put on charts (3)
- Title
- Variable labels
- Percents/quantities
Pie Chart
part of a whole
Frequency Bar Chart
- Categorical Data (its frequency)
Segmented Bar Chart
- Categorical Data
- 2 variable
- compare proportions between categorical variables
Histogram
- Quantitative Data
1 Category Displays
- pie chart
- relative bar chart
- frequency bar chart
2 Category Displays
- segmented bar
- mosaic plot
- side-by-side barchart
Two Way Table
- Marginal proportion
- Joint proportion (intersection)
- Conditional proportion (union)
- Total of 1 var and total
- 1 var cell and 2 var cell intersect (AND)
- 1 cell based on another var
Categorical variable
descriptive (words)
Ex. what school do you go to
Quantitative variable
uses numbers
Ex. how many A’s do you have
How to describe misleading graph?
(2 steps)
- what’s misleading?
- in the context of the problem
What is this question about:
“Does knowing the value of 1 var help predict the value of another var?”
Whether there’s an association btwn variables or not
Association or not?
Use 1 var to predict another
Example
- association = conditional rel frequency of 1 var is different from another
SOCS
Describing a graph
Shape
Outlier
Center
Spread
Centers (+ when to use)
Median - skewed (resistant)
Mean - symmetrical (non-resistant)
Quantitative
discrete
countable - whole num
Quantitative
continuous
measurement - decimals
Stemplot
stem + leaves
- can use 2 variables - remember one would be mirrored
- key examples: 4|5 = 45 (objects), 0|5 = 0.5 inches
Histogram ON CALCULATOR
- Make list (list > edit)
- 2nd + stat plot (y =)
- Choose plot (right list, freq 1)
- Zoom > 9 (Stat)
- Trace
Location of Median (formula)
(n+1)/2
5 Number Summary
- Minimum
- Q1
- Median
- Q2
- Maximum
In short responses, ALWAYS explain how trait affects the context of the graph/problem!!!!
please
Empirical Rule
68 - 95 - 99.7
Relationshp btwn Variance and SD
Variance = SD^2