Standardizing
What does a z-score tell us?
How many standard deviations a value is from its mean.
What is the horizontal line on the boxplot?
The median
For boxplots/5 number summaries to test for outliers.
What plot do you use when comparing groups?
Boxplot, display overally summary info better, hide details so not too much clutter.
When is election day?
November 3rd
What is the z-score formula?
z =(x-xbar)/s
z = obs-mean/standard deviation
What are the 5 numbers in the 5 number summary?
Minimum, quartile 1, median, quartile 3, maximum
What should you do when you have a confirmed outlier?
Look at the data. May be a data entry error, or just different.
When comparing groups, what aspects of the distribution should you focus on?
Shape, center, spread
Name 5 plots we can use to describe data.
Boxplot, histogram, dotplot, pie chart, bar chart, mosaic plot, scatterplot.
If we look at basketball games, and the average points scored in a game is 100 with standard deviation (SD) of 8, and we are given a game with a standardized score of 2, what was the score?
2 = (x-100)/8
16 = x -100
116
What is the formula to test for outliers on a boxplot?
If more than 1.5(IQR) + Q3 or Q1 -1.5(IQR) .
What can you learn from outliers?
Can be an extraordinary case, you can learn more from this than from the whole dataset.
If at least one distribution is skewed, what measures of center and spread should we use?
Median and IQR
Who is going to win the NBA finals this year?
Lakers (no other answers are correct sorry)
Let's say we look at the test scores of two classes.The mean of class 1 is 76 with a SD of 4 and the mean of class 2 is 81 with a SD of 8. If Carl from class 1 scores 87 and Doris from class 2 scores a 90, whose score is more unusual?
Carl: (87-76)/4 = 2.75
Doris: (90-81)/8 = 1.125
Carl's score is more unusual
How can you tell if there is skewness from a boxplot?
If the median is not in the center of the box. If the median is up(to the right) it is skewed left, if the median is down (to the left) it is skewed right. Also look at lengths of whiskers.
If we ask students in a school "How many languages do you speak?" and we get the following information: Q1: .5 and Q3: 1.5. We see on our plot a point at 10. Is this an outlier, and is it feasible?
1.5(1) = 1.5 Q3 + 1.5 = 3, yes an outlier, not feasible who speaks 10 languages?
What happens if you are comparing two histograms with two different scales?
Should you take a move-in cart to the beach with 500 other students, party and then post the video to social media, all during a global pandemic?
Heck no. Please be smart.
When we standardize what happens to:
the shape?
the mean?
Shape remains the same, mean becomes 0, sd becomes 1.
Let's say we have the following 5 number summary:
(10,15,22,40,56)
If we have an observation at 4 is that an outlier? What about 76?
IQR = 25, 1.5x(25) = 37.5
Q3 + IQR = 77.5
Q1 - IQR = -22.5
Neither is an outlier
Which plot is easiest for viewing outliers?
Boxplot. Technology will have points on either sides of max/min if an outlier. Histograms do not highlight outliers like this.
Can you determine modality, mean or SD from a boxplot? If so, how?
Cannot
Which professor narrated the lecture videos 3-4 and 3-5 (Monday and Tuesday's videos)?