Guide to Using Minitab For Basic Statistical Applications

Download Report

Transcript Guide to Using Minitab For Basic Statistical Applications

Guide to Using Minitab For Basic
Statistical Applications
To Accompany
Business Statistics: A Decision Making
Approach, 6th Ed.
Chapter 3:
Describing Data Using Numerical
Measures
By
Groebner, Shannon, Fry, & Smith
Prentice-Hall Publishing Company
Copyright, 2005
Chapter 3 Minitab
Examples

Population Mean
Foster City Hotel
 Mean, Median,and Mode
Weigh-In-Motion
 Percentiles and Quartiles
Weigh-In-Motion
 Box and Whisker Plots
Weigh-In Motion
More Examples
Chapter 3 Minitab Examples (continued)


Measures of Variation
Weigh-In-Motion
Empirical Rule
Burger N’ Brew
Population Mean
Foster City Hotel
Issue:
Determine the mean nightly revenue for
the Foster City Hotel.
Objective:
Use Minitab to calculate the population
mean revenue Data File is FosterCity.mtw
Population Mean – Foster City Hotel
Open the Minitab file
called FosterCity.mtw
Population Mean – Foster City Hotel
Select Calc – Column
Statistics
Population Mean – Foster City Hotel
Click on Mean and
define the Input
Variable for
analysis
Population Mean – Foster City Hotel
Population Mean
Population Mean – Foster City Hotel
Second method: Select
Stat, then Basic
Statistics then Display
Descriptive Statistics
Population Mean – Foster City Hotel
Select the variables
to analyze
Population Mean – Foster City Hotel
Population Means
Note, other output
is also provided
Mean, Median and Mode
Weigh-In-Motion
Issue:
Does the WIM scale produce gross
weights that are close to the POE weights?
Objective:
Use Minitab to develop histograms for the
weights from each scale and to compute the
mean, median, and mode for each scale. The
data file is Trucks.mtw.
Mean, Median, and Mode – Weigh-In-Motion
Open the data
file called
Trucks.mtw
Data File contains
200 trucks.
Mean, Median, and Mode – Weigh-In-Motion
Construct cutpoints column.
Mean, Median, and Mode – Weigh-In-Motion
Select Graph – Histogram to
form the histograms for the
POE and WIM gross weights
Mean, Median, and Mode – Weigh-In-Motion
Identify variables to be
graphed and set Display to
Bar
Next, use Options to set
class widths
Mean, Median, and Mode – Weigh-In-Motion
For Type of Intervals select
Cutpoint, in Definitioin of
intervals, insert Cutpoint
column.
Mean, Median, and Mode – Weigh-In-Motion
Use Edit Attributes to Set
Graph Colors, etc.
Mean, Median, and Mode – Weigh-In-Motion
Use Annotate to Set Graph
Titles, etc.
Mean, Median, and Mode – Weigh-In-Motion
To modify X axis labeling,
use Frame - Tic
Mean, Median, and Mode – Weigh-In-Motion
POE Gross Weight
Histogram
Completed
Mean, Median, and Mode – Weigh-In-Motion
Repeat the process for WIM
gross weights. Then
compute mean, median and
modes for both variables
Mean, Median, and Mode – Weigh-In-Motion
Select Stat – then select
Basic Statistics and Display
Descriptive Statistics
Mean, Median, and Mode – Weigh-In-Motion
Select the variables to be
analyzed
Mean, Median, and Mode – Weigh-In-Motion
Mean and Median
Minitab does not
compute the mode
Percentiles and Quartiles
Weigh-In-Motion
Issue:
Determine Percentiles and Quartiles for
WIM and POE Gross Weights
Objective:
Use Minitab to calculate 10th percentiles
and 1st and 3rd quartiles for weigh-in-motion
data. Data file is Trucks.mtw
Percentiles and Quartiles - Weigh-In-Motion
Open the data
file called
Trucks.mtw
Data File contains
200 trucks.
Percentiles and Quartiles - Weigh-In-Motion
Select Stat – then Basic
Statistics – then Display
Descriptive Statistics
Percentiles and Quartiles - Weigh-In-Motion
Define Variables for
analysis
Percentiles and Quartiles - Weigh-In-Motion
1st and 3rd quartiles
for WIM and POE
gross weights
Percentiles and Quartiles - Weigh-In-Motion
To obtain percentiles, Select Stat
– Reliability/Survival –Parametric
Dist Analysis Right Cens…
Percentiles and Quartiles - Weigh-In-Motion
Identify Variables for
analysis – then click
on Estimate
Percentiles and Quartiles - Weigh-In-Motion
Enter desired
percentiles
Percentiles and Quartiles - Weigh-In-Motion
1st and 3rd quartiles
WIM Gross Weight
WIM Gross Weight
10th Percentile
Percentiles and Quartiles - Weigh-In-Motion
1st and 3rd quartiles
POE Gross Weight
POE Gross Weight
10th Percentile
Box and Whisker Plots
Weigh-In-Motion
Issue:
Analyze how WIM and POE gross weights
compare.
Objective:
Use Minitab to develop a Box and Whisker
Plot for comparing WIM and POE gross weights.
Data file is Trucks.mtw
Box and Whisker Plots - Weigh-In-Motion
Open the data
file called
Trucks.mtw
Data File contains
200 trucks.
Box and Whisker Plots - Weigh-In-Motion
Click on Graph –
then select
Character Graphs –
then select Boxplot
Box and Whisker Plots - Weigh-In-Motion
Enter Variable for
Analysis (WIM Gross
Weight)
Box and Whisker Plots - Weigh-In-Motion
Box and Whisker
Plot – WIM Gross
Weight
Box and Whisker Plots - Weigh-In-Motion
Repeat process for
Box and Whisker
Plot – POE Gross
Weight
Measures of Variation
Weigh-In-Motion
Issue:
Understand the Variation in POE and WIM
gross Weights
Objective:
Use the Minitab to compute various
measures of variation in WIM and POE Gross
Weights Data file is Trucks.mtw
Measures of Variation - Weigh-In-Motion
Open the data
file called
Trucks.xls
Data File contains
200 trucks.
Measures of Variation - Weigh-In-Motion
Click on Stat tab – then select
Basic Statistics – then choose
Display Descriptive Statistics
Measures of Variation - Weigh-In-Motion
Define variables to be
included in the analysis
Measures of Variation - Weigh-In-Motion
Standard Deviation,
Minimum and Maximum
values
Empirical Rule
Burger N’ Brew
Issue:
Analyze the Phoenix Burger Sales
Distribution
Objective:
Use the Minitab to compute graphs and
numerical measures necessary for using the
empirical rule to analyze sales at Burger N’ Brew
Data file is BurgerNBrew.mtw
Empirical Rule – Burger N’ Brew
Open the data file
called
BurgerNBrew.mtw
Data File contains
sales for 365 days.
Empirical Rule – Burger N’ Brew
This tutorial will
demonstrate a
different way to
develop a Histogram
using class widths of
2. Select Graph Histogram
Empirical Rule – Burger N’ Brew
Identify variable for
analysis – Then use
Options to specify
classes
Empirical Rule – Burger N’ Brew
Use class widths of
size 2
Empirical Rule – Burger N’ Brew
Use Annotate to
specify titles etc.
Empirical Rule – Burger N’ Brew
Finished Histogram –
Bell Shaped Distribution
Empirical Rule – Burger N’ Brew
To compute numerical measures –
Click Stat – Basic Statistics – Display
Descriptive Statistics
Empirical Rule – Burger N’ Brew
Identify variable for
analysis
Empirical Rule – Burger N’ Brew
Mean and Standard Deviation
Empirical Rule:
68% within 15.12 + (1)3.13
95% within 15.12 + (2)3.13
99.7% within 15.12 + (3)3.13)