HR Analytics Applications of Comparison of Means and ANOVA

Chapter 8
HR Analytics Applications of
Comparison of Means and ANOVA
Learning Objectives
• To understand the application of t-test and analysis of

variance (ANOVA)
• To interpret output using both Analysis ToolPak and R GUI
One-Sample T-test
• The one-sample t-test tells us if a sample mean comes from

a population with a definite mean.
• For this variant of t-test, the alternative hypothesis
assumes that there exists a difference between the specific
value and the sample mean.
• The null, on the contrary, assumes that there exists no
difference between this specific value and the sample
value. Thus,
• H0: sample mean = specified value
• H1: sample mean ≠ specified value
Conducting One-Sample T-test in PSPP
Conducting One-Sample T-test in RCMDR
Assumptions of One-Sample T-test
• The data must be continuous.

• The sample observations are independent of one another.
• Similarly, outliers should be checked. The scores must be
standardized and values outside the range of –3.29 and
3.29 are removed.
• The data should be approximately normally distributed.
Since real world data is not perfectly normal, statisticians
have optimized the situation to be approximately normal.
Whether a data set is normal or not can be checked using
the histogram and Kolmogorov–Smirnov Test.
Paired Sample T-test
• This test is also known by the names of dependent t-test

and repeated measures t-test.
• In this variant of the t-test, same unit is measured twice
resulting in pairs of values.
• For this variant of t-test, the alternative hypothesis
assumes that there exists a difference between the mean
values of the pairs. The null, on the contrary, assumes that
there exists no difference between the means of the pairs.
Thus,
• H0: Difference of means of pairs = 0
• H1: Difference of means of pairs ≠ 0
Conducting T-test in RCMDR
• Step 1: Insert Excel file

• Step 2: Specify the group for which the test has to be done
using the Manage variables in active data set
• Step 3: Click on convert numeric variables into factors
• Step 4: Specify the levels
• Step 5: Go to Statistics menu
• Step 6: Click on Means sub-menu and select ‘Independent
Sample T-test’
• Step 7: Specify the response variable
• Step 8: Specify whether equal variance is assumed or not
Conducting Paired Sample T-test in RCMDR
Independent Sample T-test
Features:
• One independent variable
• Two groups, or levels of the independent variable
• Independent samples (between groups): the two groups are not
related in any way
• Interval/ratio measurement of the dependent variable
Basic Concepts:
• The t-stat is determined based on the level of significance and
the sample size (degrees of freedom)
• As degrees of freedom increase, the t-distribution becomes
increasingly normal.
• Degree of freedom = no. of variables allowed to vary
Conducting Independent Sample T-test in PSPP
Conducting Independent sample T-test in
RCMDR
ANOVA
• Technique to compare difference in mean
output/attitude/performance (continuous variable) between
three or more groups of employees/product lines/sales
teams/students.
• When there is only one dependent variable.
• Not meant for categorical dependent variable.
• Assumption: Groups should be homogeneous (indicated by
Levene’s test; should be insignificant).
Types of ANOVA
• One way—the groups are different in terms of only one
level/factor.
• Example: comparing marks of MBA, BBA and law students
• Two way—the groups are different in terms of two
levels/factors.
• Example: comparing training performance of executives (male
versus female) using two different training techniques (OJT
versus simulation)
One-Way ANOVA
• Evaluate the difference among the means of three or
more groups.
Examples: Accident rates for 1st, 2nd, and 3rd shift.
Expected mileage for five brands of tires.
• Assumptions
• Populations are normally distributed.
• Populations have equal variances.
• Samples are randomly and independently drawn.
Conducting One-Way ANOVA in PSPP
Conducting ANOVA in RCMDR
• Step 5: Go to Statistics menu
• Step 6: Click on Means sub-menu and select ‘One-Way ANOVA’
• Step 7: Specify the grouping variable and the response variable
• Step 8: Select ‘Pair-wise comparison’
Conducting One-Way ANOVA in RCMDR (Continued)
Assumptions for Experimental Designs
• The groups are homogeneous (equal variance).
• The groups are subjected to the treatment randomly (random
allocation of treatment).
• The treatments are not given to the groups with the knowledge
of the experimenter (allocation concealment).
• The treatments are considered to be unique by the
experimental participants.
ANOVA-test for comparing more groups of data
• T-test—to compare means of two groups (e.g. comparing mean
performance of two teams)
• ANOVA—to compare means of more than two groups (e.g. to
compare the output of three different production plants)
• Both are used to see whether the groups are significantly
different in terms of their mean values/outputs from each other
• ANOVA = Multiple t-tests
• Then why do it?
Reason for Doing ANOVA Instead of Multiple T-
tests
• Type I error—the error committed by researchers by rejecting a
null hypothesis when it is actually true.
• In simple terms this means—to think there is an effect of IV on
DV when actually there is none; or to think there is a correlation
when actually the variables are uncorrelated.
• Type II error—It is the reverse of Type I error—to accept a null
hypothesis when it is actually false.
• That is, failing to detect an effect when there is one.
Why do Type I, II Errors Occur?
• Main reason for such errors is human mistake.
• To allow for such mistakes, statisticians resort to type I error
rate, measured on the basis of confidence interval.
• Confidence interval signifies the extent to which a researcher is
sure that her/his results are correctly interpreted.
• Generally set at 95 per cent, which means that out of 100 times,
the researcher may be wrong in her/his interpretation 5 times.
• Therefore, type I error rate for one test = (1 – 0.95) = 0.05.
Why ANOVA Is Better than T-test?
• One t-test causes a type I error of [1 – 0.95] = 0.05.
• For a t-test to compare three groups, three separate two group
comparisons have to be done.
• Therefore total error = [1 – (.95)  3] = 1 – 0.857 = 0.143 =
14.3%
• Thus, error rate has increased from 5 per cent to 14.3 per cent.
Conducting T-test in RCMDR
• Step 5: Go to Statistics Menu
• Step 6: Click on Means sub-menu and select “Independent
sample t-test”
• Step 7: Specify the response variable
• Step 8: Specify whether equal variance is assumed or not
THANK YOU

HR Analytics Applications of Comparison of Means and ANOVA

Uploaded by

Copyright:

Available Formats

HR Analytics Applications of Comparison of Means and ANOVA

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HR Analytics Applications of Comparison of Means and ANOVA

Uploaded by

Copyright:

Available Formats

Chapter 8

• To understand the application of t-test and analysis of

• The one-sample t-test tells us if a sample mean comes from

• The data must be continuous.

• This test is also known by the names of dependent t-test

• Step 1: Insert Excel file

You might also like