stats 250 lab 2 julie ghekas [email protected] september 15, 2014

44
STATS 250 Lab 2 Julie Ghekas [email protected] September 15, 2014

Upload: mark-henry

Post on 26-Dec-2015

222 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

STATS 250 Lab 2Julie Ghekas

[email protected]

September 15, 2014

Page 2: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Schedule

• Recap from first lab and prelab

• Warm Up

• Lab 2

• Cool Down

• iClicker Questions

Page 3: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Lab Workbook for Fall 2014

• open.umich.edu/

– Search Stats 250

– Select Materials Tab for the course

– Scroll down to have access to handouts, labs, lecture notes,

etc.

• open.umich.edu/education/lsa/statistics250/fall2015/

materials#Labs

• Or order from Amazon (~$10)

• Recommend taking notes in updated personal workbook

Page 4: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Prelab Results

Right skewed: Mean > Median

Left skewed: Median > Mean

Symmetric: Mean = Median

Remember, the mean is more sensitive to outliers than the median.

Page 5: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Recap for Homework

• Practice homework graded to the

same standard of regular homework

• Answer the question fully

• Put your name on generated graphs

• Show all work

• Include units

Page 6: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Boxplots

• Graph of 5-number summary

• Outliers denoted with ° and *

• Can be side-by-side

• Does not show shape of distribution

Page 7: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Bar charts

• Displays categorical variables

• Y-axis represents counts, proportions,

or percentages

• Can rearrange bars in any order

Page 8: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 9: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 10: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Time Series/Sequence Plots

• Examining data over time

• Checks assumption that observations

are from an identically distributed

population

• Be careful; time series can be

displayed in different formats

Page 11: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Time Series

Source: http://www.statcan.gc.ca/edu/power-pouvoir/ch9/bargraph-diagrammeabarres/5214818-eng.htm

Source: http://blogs.bgsu.edu/statgraphicsmepaler/2013/03/22/new-havens-temperature-in-4-different-time-series-plots/

Page 12: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Time Series: Trends• A trend is a consistent, long-term rise

or fall.

Page 13: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Time Series: Variation

• Generally, variation is used to

describe patterns in the data.

Seasonal VariationIncreasing Variation

Page 14: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Time Series: Stability

• If there are no patterns in the time plot, then

it is considered stable.

• Stability helps us confirm or reject the

identically distributed part of iid/random

sample. In order for data to be considered

stable, both the mean and the variance of the

observations needs to be constant over time.

Page 15: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Q-Q Plots

• Checks assumption that observations are

from a normally distributed parent population

• Q stand for quantiles (percentiles): graph

compares Quantiles from the standard normal

distribution with Quantiles from our sample

• Want a straight line

• Better than a histogram

Page 16: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Q-Q Plot of Data from an Approximately Normal Distribution

Page 17: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Q-Q Plots that do NOT allow us to assume a population with Normal Distribution

Page 18: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

R scripts

• Canvas homepage -> R tutorials

• Open timeseries.rdata or qqplot.rdata

– Canvas homepage -> R tutorials ->Time

Series/QQ plots

• Start script with timeseries() or qqplot()

– Without the underscore printed in the lab

workbook

Page 19: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Warm Up

Page 20: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Lab

• With a partner or two, work on the Lab

• You will not get credit if you work alone

• Work with employee data.sav

• If you finish early, complete the Cool

Down, R practice, Example Exam

Question, or Practice HW problem 3

Page 21: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 22: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 23: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 24: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

IQR=$13,162.50

IQR=$7,125.00

IQR=$16,200.00

Page 25: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 26: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 27: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Cool Down

• Everyone turns in own ticket

• Work on Cool Down in groups

Page 28: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Survey: Students were asked how many hours they study in

a typical week. A five-number summary of the responses is:

2, 10, 14, 20, 60

Fill in the blank: About 75% of the students spent at least

___ hours studying in a typical week.

A. 10

B. 14

C. 20

D. 45

Page 29: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Survey: Students were asked how many hours they study in

a typical week. A five-number summary of the responses is:

2, 10, 14, 20, 60

Fill in the blank: About 75% of the students spent at least

___ hours studying in a typical week.

A. 10

B. 14

C. 20

D. 45

Page 30: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Survey: Students were asked how many hours they study in

a typical week. A five-number summary of the responses is:

2, 10, 14, 20, 60

What percent of students reported studying between 10

and 20 hours in a typical week?

A. 68%

B. 50%

C. 25%

D. 75%

Page 31: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Survey: Students were asked how many hours they study in

a typical week. A five-number summary of the responses is:

2, 10, 14, 20, 60

What percent of students reported studying between 10

and 20 hours in a typical week?

A. 68%

B. 50%

C. 25%

D. 75%

Page 32: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Which of the following provides the most

information about the shape of a data set?

A. Boxplot

B. Pie chart

C. Five number summary

D.Histogram

Page 33: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Which of the following provides the most

information about the shape of a data set?

A. Boxplot

B. Pie chart

C. Five number summary

D.Histogram

Page 34: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClickerHere is a graph showing revenue for a

company. What kind of graph is this?

• A. Bar chart

• B. Histogram

• C. Time plot

• D. Box plot

0

5

10

15

20

25

Re

ve

nu

es

($

billi

on

s)

Actual Revenue for Eastman Kodak

Page 35: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClickerHere is a graph showing revenue for a

company. What kind of graph is this?

• A. Bar chart

• B. Histogram

• C. Time plot

• D. Box plot

0

5

10

15

20

25

Re

ve

nu

es

($

billi

on

s)

Actual Revenue for Eastman Kodak

Page 36: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClickerCost of a gallon of gas

Can we describe this graph as…

A. Right Skewed B. Left Skewed C.

Neither

Page 37: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClickerCost of a gallon of gas

Can we describe this graph as…

A. Right Skewed B. Left Skewed C.

Neither

Page 38: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

In a boxplot, what does a dot represent?

A. Quartile

B. Mean

C. Median

D. Mode

E. Outlier

Page 39: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

In a boxplot, what does a dot represent?

A. Quartile

B. Mean

C. Median

D. Mode

E. Outlier

Page 40: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Which time plot is the most stable?

A.

B.

C.

Page 41: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

Which time plot is the most stable?

A.

B.

C.

Page 42: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

iClicker

How do you feel about the material covered today?

A. Completely understood everything

B. Understood main ideas, shaky on details

C. Good for the first half, lost for the second

D. Had trouble with some main ideas

E. Difficulty following most materials

Page 43: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014
Page 44: STATS 250 Lab 2 Julie Ghekas jghekas@umich.edu September 15, 2014

Reminders• Good job setting up LectureBook

• Practice Homework due Thursday 8

am

• Pre-lab 3 due Monday 8 am

• Office Hours

• Food Allergies?