Test On Variables

Group 5
Asim Kumar Verma
Deeptiman Guha
Himanshu Arora
Kaavish Kidwai
Nirmal Modh

In Surveys, the foolish ask questions, wise cannot answers .

-Oscar Wilde
 Variable
 Individual: The objects described by a set
of data, individuals may be people,
animals or things
 Ex: Students in an AP Statistics class
 Variable: any characteristic of an
individual. A variable can take on
different values for different individuals
 Ex: Heights of individuals: Student #1: 5’ 5”
Student #2: 5’ 7” Student #3: 5’ 2”
 Two Types of Variables
 Quantitative:
 Numeric value that it makes sense to do arithmetic
operations (+, -, x, /)
 Finding an arithmetic “average” makes sense
 EX:
 Height
 Weight
 Age
 Income
 Test Scores
 Types of Variables
 Categorical
 Records which of several groups or categories
to which an individual belongs
 If there are only 2 possible categories, the
variable is called binary
 Ex:
 Gender
 Race
 Marital status
Steps in Hypothesis Testing
1. Hypothesis formulation
2. Specify the significance level
3. Collect the data and calculate the test
statistic
4. Fetch the tabulated value
5. Make the statistical decision
 Statistical Test

Parametric Non Parametric

When n
When n
less than
greater than 30
30
 Hypothesis testing
 Goal: Make statement(s) regarding unknown population
parameter values based on sample data
 Elements of a hypothesis test:
 Null hypothesis - Statement regarding the value(s) of
unknown parameter(s). Typically will imply no association
between explanatory and response variables in our applications
 Alternative hypothesis - Statement contradictory to the null
hypothesis (will always contain an inequality)
 Test statistic - Quantity based on sample data and null
hypothesis used to test between null and alternative hypothesis
 Rejection region - Values of the test statistic for which we
reject the null in favor of the alternative hypothesis
 Test On Variables
In these decisions are made on quantitative measures
 T test

Statistical hypothesis test in which the test statistic

follows a student's t distribution if the null
hypothesis is true n < 30
 Z Test

A Z-test is any statistical test for which the

distribution of the test statistic under the null
hypothesis can be approximated by a normal n > 30
 Hypothesis Testing: Steps

Test the Assumption that the true mean

grade point average of juniors is at least 3.
 1. State H0 H0 : m ³ 3.0
 2. State H1 H1 : m < 3.0
 3. Choose a a = .05
 4. Choose n n = 100
 5. Choose Test: Z test
 Hypothesis Testing: Steps
(continued)

Test the Assumption that grade point average of

juniors is at least 3.

 6. Set Up Critical Value(s) Ze= 1.645

 7. Collect Data 100 students sampled
 8. Compute Test Statistic Computed Test Stat.= -2
 9. Make Statistical Decision Since Z mod greater than Ze
Reject Null Hypothesis
 10. Express Decision The true mean grade point is less
than 3.0
 Rejection Region

H0: m ³ 0 H0: m £ 0
H1: m < 0 H1: m > 0
Reject H0 Reject H 0

a a

0 Z 0 Z
Must Be Significantly Small values don’t contradict H0
Below m = 0 Don’t Reject H0!
 Hypothesis Test
• Hypothesis testing is used when we want to
know whether the sample data support some
preconceived theory we hold about the
parameters of the population model.
• If we can find a numerical value of the
relevant probability, then we might be able to
assess whether the belief is supported or not:
• a high probability of obtaining the data
supports the belief, a low one does not.
 t test
• t tests is simply comparing two means to see if
they are significantly different from each other.
• The more technical definition or description of a
t test is any statistical test that uses the t, or
Student's t, family of distributions.

Most Economists think of god as working great multiple

regressions in the sky – Edgar Fiedler
 Student’s t-test
If random samples of size less than 30 are taken from a normal distribution and the
samples used to estimate the variance, then the statistic
ỹ-µ
−−−−
s/√n
is not normally distributed. The probabilities in the tails of this distribution are
greater than for the standard normal distribution
This is reasonable since
ỹ-µ
z= −−−−−
s/√n
contains only one random variable ỹ, while
ỹ-µ
−−−−−
s/√n
contains two random variables y and s. As n increases this new distribution
approaches the standard normal distribution.

ỹ-µ
 Properties of t Test
Student’s t distributions are
1. Unimodal;
2. Asymptotic to the horizontal axis;
3. Symmetrical about zero, E(t);
4. Dependent on v, the degrees of freedom (for the
statistic under discussion, v=n-1);
5. More variable than the standard normal
distribution, V(t)=v/(v-2) for n > 2;
6. Approximately standard normal if v is large.
 Excel Command
 To test whether the population mean could be
µ, use :
 MeanTest uses the original data and the
hypothetical value µ of the population mean.
 StudentTPValue uses only the value of the test
statistic t = m-µ/Sm and degrees of freedom n - 1.
 The two commands use the Student t distribution to
perform the testing.
 Comparision

Comparision of standard normal distribution and a t

distribution
 Standard Error

 Test determines whether differences in two

samples is significant enough to suggest a
difference in respective populations.

 Hence average expected difference between the

means of two samples is required. This average
difference is known as the Standard Error of the
difference between two means.
 Paired Samples
 A dependent samples t test is also used to compare two
means on a single dependent variable.
 Unlike the independent samples test, however, a
dependent samples t test is used to compare the means
of a single sample or of two matched or paired samples.
 For example, if a group of students took a math test in
March and that same group of students took the same
math test two months later in May, we could compare
their average scores on the two test dates using a
dependent samples t test.
 Example
 Suppose we want to know whether employees at our
widget-making factory are more productive after they
return from a 2-week vacation.
 We randomly select 30 of my employees and calculate the
average number of widgets made by each employee
during the week before they go on vacation.
 We find that, on average, the employees made 250
widgets each during the week. During the week after
they return from vacation, I keep track of how many
widgets is made by the same sample of 30 employees and
find that, on average, they made 300 widgets each during
the week after returning from their vacations.
 Example (contd.)
 The crux is whether what we observed in this sample
represents a likely difference in the productivity of
the larger population of widget makers after a
vacation.
 distinction between this dependent samples t test
and the independent samples t test is that rather than
comparing two samples on a single dependent
variable, now we are comparing the average scores
of a single sample (i.e., the same group of 30
employees) on two variables.
 Contd…
 T=observed diff. between prevacation and postvacation
means/standard error of the difference between the
means
Or
T=x-y/Sd

 The formula for calculating the standard error of the

difference between the means for dependent samples is
slightly different than the one for independent samples, but
the principles involved (i.e., what the standard error
represents) are the same.
 Case
 Effectiveness of Newly Implemented
“Green Belt Project” in a Call Center of
Insurance Company
 Bibliography
 Paradigm : IMT G Journal
 Applied statistics in business and economics by
Doane & Seward
 www.wikipedia.com
 Case: An analysis of the Indian Textile Industry
in Quota Free Regime by Manisha Sharma &
Anu Prashaant

It is not nice to be wedded to anything – not even a

theory – Samuel Butler