1. SIT DOLOR AMET
Studentโs t
Distribution
& Its
Applications
Savitri Dasigi
Paul Paulson
Vidit Jain
Prof.Kavya S
Presented by: Mentor:
2. t - distribution
The t-distribution (also called Studentโs t distribution) is
a family of distributions that look almost identical to the
normal distribution curve, only a bit shorter and fatter.
The t distribution is used instead of the normal
distribution when you have small samples (for more on
this, see: t-score vs. z-score). The larger the sample size,
the more the t distribution looks like the normal
distribution.
3. History
โข Just past the turn of the 19th century, a major development in the science was
fermenting at Guinness Brewery.
โข 1908 - William Sealy Gosset faces a problem at the Guinness brewery
where he worked.
โข There didn't exist a better way to test samples at the time for accuracy of
representing the population and measure accuracy of their estimates.
โข t-distribution thus arises - a sampling distribution of normal samples with
an unknown population variance.
โข It thus became a primary way to understand likely error of an estimate and
becomes source of statistical significance tests.
6. t -test
A t-test is a type of inferential statistic used to
determine if there is a significant difference
between the means of two groups, which may
be related in certain features.
The t-test is one of many tests used for the
purpose of hypothesis testing in statistics.
7. Applications of t-test
Testing
significance
between
hypothetical
mean of sample
and population
mean.
01
Testing
significance
between mean of
2 different
samples.
02
Testing
significance
between mean of
sample post &
prior to stimuli
operation.
03
Testing significance
b/w hypothetical
rank correlation
coefficient and
population rank
correlation
coefficient
04
Testing line of
regression
analysis
05
Testing
significance of
partial and
multiple
correlation
coefficient .
06
8. The One Sample t Test
determines whether the
sample mean is
statistically different from
a known or hypothesized
population mean.
ONE
SAMPLE
TEST
A two-sample t-test is
used to test the difference
(d0) between two
population means. A
common application is to
determine whether the
means are equal.
INDEPEN
DENT
TWO
SAMPLE
TEST
The Paired Samples t Test
compares two means that
are from the same
individual, object, or
related units.
PAIRED
SAMPLE
TEST
Types of t test
9. Assumptions
โขThe scale of measurement applied to the data collected must follow a
continuous or ordinal scale.
โขThe data to be tested should be a simple random sample, (i.e.) the data
should be collected from a representative, randomly selected portion of
the total population.
โขThe population data must follow the normal distribution
โขHomogeneity of variance is assumed and population variance is
unknown.
10. Test Procedure for Two Sample t-test
1.Defining the Hypothesis
โข ๐ฏ๐ : ๐๐ = ๐๐ (Independent population means are same)
โข ๐ฏ๐ : ๐๐ > ๐๐ ( right tailed)
Or
โข ๐ฏ๐ : ๐๐ < ๐๐ ( left tailed)
0r
โข ๐ฏ๐ : ๐๐ โ ๐๐ (2 tailed)
11. The test statistic is a t statistic (t) defined by the following equation:
2. Test statistic.
Where, ๐ฅ =
๐ฅ
๐1
๐ฆ =
๐ฆ
๐2
๐ 2
=
1
๐1+๐2โ2
( (๐ฅ๐ โ ๐ฅ )2
+ (๐ฆ๐ โ ๐ฆ )2
๐1 ๐๐๐ ๐2 ๐๐๐ ๐กโ๐ ๐ ๐๐๐๐๐ ๐ ๐๐ง๐๐
๐ก =
๐โ๐
๐
๐
๐๐
+
๐
๐๐
~students t distribution with (n1 +
n2 โ 2 )degrees of freedom
12. 3.Critical region and conclusion
(i) If ๐ป1: ๐1>๐2 ,then the critical value will be ๐ก๐ผ(๐1+๐2 โ 2)
we reject ๐ป0 if ๐ก๐๐๐ > ๐ก๐ผ(๐1+๐2 โ 2) or P ( T > t ) โค ๐ผ
(ii) If ๐ป1: ๐1 < ๐2 ,then the critical value will be โ๐ก๐ผ(๐1+๐2 โ 2)
we reject ๐ป0 if ๐ก๐๐๐ < - ๐ก๐ผ(๐1+๐2 โ 2) or P ( T < -t ) โค ๐ผ
13. iii) If ๐ป1: ๐1 โ ๐2 ,then the critical value will be ๐ก๐ผ/2(๐1+๐2 โ 2) and โ๐ก๐ผ/2(๐1+๐2 โ 2)
we reject ๐ป0 if |๐ก๐๐๐| > ๐ก๐ผ/2 (๐1+๐2 โ 2)
Or if ๐ก๐๐๐ is falling outside the region (-๐ก๐ผ/2 , ๐ก๐ผ/2 ) or P ( T >|t| ) โค ๐ผ
14. Test Procedure of Paired t test
1.Defining the Hypothesis
โข ๐ฏ๐ : ๐๐ = ๐๐ (Two related population means are same)
โข ๐ฏ๐ : ๐๐ > ๐๐ ( right tailed)
Or
โข ๐ฏ๐ : ๐๐ < ๐๐ ( left tailed)
0r
โข ๐ฏ๐ : ๐๐ โ ๐๐ (2 tailed)
15. 2. Test Statistic
The test statistic for the paired Samples t test, denoted t, is given by :
Where, d=๐ฅ๐ - ๐ฆ๐
๐ = ๐=1
๐
๐๐
๐
s=
1
๐โ1 ๐=1
๐
(๐๐ โ ๐)2
t =
๐
๐
๐
~ students t distribution with (n-1) degrees of freedom
16. 3.Critical region and conclusion
(i) If ๐ป1: ๐๐ฅ>๐๐ฆ ,then the critical value will be ๐ก๐ผ(๐ โ 1)
we reject ๐ป0 if ๐ก๐๐๐ >๐ก๐ผ(n โ 1 ) or P ( T > t ) โค ๐ผ
(ii) If ๐ป1: ๐๐ฅ < ๐๐ฆ ,then the critical value will be โ๐ก๐ผ(n - 1)
we reject ๐ป0 if ๐ก๐๐๐ < - ๐ก๐ผ(๐ โ 1) or P ( T < -t ) โค ๐ผ
17. iii) If ๐ป1: ๐๐ฅ โ ๐๐ฆ ,then the critical value will be ๐ก๐ผ/2(n -1 ) and โ๐ก๐ผ
2
(๐ โ 1)
we reject ๐ป0 if| ๐ก๐๐๐| > ๐ก๐ผ/2 (n - 1)
Or if tcal is falling outside the region (-๐ก๐ผ/2 , ๐ก๐ผ/2 ) or P ( T > |t| ) โค ๐ผ
18. Introduction
And this
proved to be
the best
methodology
for
understating
and predicting
human
behavior
Since the
dawn of the
Sciences of
Psychology
and Statistics
,both had
worked hand
in hand
And with the
breakthrough in
the past many
cooperate
sectors use
psychometric
assessments
This Project
presents a basic
model ----------
big 5
psychometric test
paired with
sampling survey
and Statistical
tools like T-test.
20. Big Five
Personality
Test
The big five personality are the most commonly used model of
personality in academic psychology.
The five personality variables which stand out in explaining a
personality are :
Extraversion,neuroticism,aggreableness,conscientiouness and
openness to experience.
This test uses the Big-Five Factor Markers from the International
Personality Item Pool developed by Goldberg(1992)
21. Our Sample Survey:
โข OBJECTIVE: To compare five basic behavioural traits among the different group of
students at NMIMS, Bangalore
โข Source of Questionnaire: openpsychometrics.com
โข Method of Data Collection: Primary data through google forms
โข Total Samples collected :30
23. t- Test Assumptions Verification : Normality of Data and Random Samples :
Shapiro Wilks Test for Normality:
R Code with Sample Output:
Q-Q plot (or quantile-quantile
plot) draws the correlation
between a given sample and the
normal distribution.
R Code with Sample Output:
24. Hypothesis for Independent two sample t-test for Level of Education
Let Ho : There is no significant difference in the mean scores of UG and PG students (i.e.) ๐1 = ๐2
H1 : There is a significant difference in the mean scores of UG and PG students (i.e.) ๐1 โ ๐2
The above general hypothesis is applied for all five domains namely Extroversion, Agreeableness,
Consciousness, Emotional Stability and Imagination and the analysis is done .
27. Conclusions
โข Extroversion:
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Extroversion between the students of UG and PG.
โข Agreeableness
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Agreeableness between the students of UG and PG.
โข Consciousness
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Consciousness between the students of UG and PG.
โข Emotional Stability
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Emotional Stability between the students of UG and PG.
โข Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Imagination between the students of UG and PG.
28. Hypothesis for Independent t-test taking for Gender
Let Ho : There is no significant difference in the mean scores of Male and female students (i.e.) ๐1 = ๐2
H1 : There is a significant difference in the mean scores of Male and Female students (i.e.) ๐1 โ ๐2
The above general hypothesis is applied for all five domains namely Extroversion, Agreeableness,
Consciousness, Emotional Stability and Imagination and the analysis is done .
29. S.No Domain Size(N) Mean S.D S.E. T value D.o.F P-value
1 Extroversion
Male 15 17.33 6.24 1.69
-2.2421 28 0.03306
Female 15 21.266 5.13 0.94
2 Agreeableness
Male 15 21.60 5.28 1.36
2.1914 28 0.03691
Female 15 18.466 1.68 0.43
3 Consciousness
Male 15 19.33 3.96 1.02
-1.386 28 0.1768
Female 15 21.400 4.21 1.09
4
Emotional
Stability
Male 15 17.0 3.21 0.83
-1.7686 28
0.08767
Female 15 19.8 5.23 1.35
5 Imagination
Male 15 21.0 3.76 0.94
0.044145 28 0.9657
Female 15 21.933 4.48 1.16
Independent two sample t-test Analysis with Gender as a Parameter:
30. Conclusions
โข Extroversion:
Since p-value < 0.05 , we reject the null hypothesis and hence we conclude that there is a significant difference
in the Extroversion between the male and female students (i.e.) female students are more extrovert than the
male students.
โข Agreeableness
Since p-value < 0.05 , we reject the null hypothesis and hence we conclude that there is a significant
difference in the Agreeableness between the male and female students.
โข Consciousness
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Consciousness between the male and female students.
โข Emotional Stability
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Emotional Stability between the male and female students.
โข Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Imagination between the male and female students.
31. Hypothesis for Paired t -test Analysis:
Let Ho : There is no significant difference between the mean scores of any two pairs of behavioural traits
(i.e.) ๐1 = ๐2
H1: There is a significant difference between the mean scores of any two pairs of behavioural traits (i.e.)
๐1 โ ๐2
This general hypothesis is applied for all the permutations of the domains namely Extroversion &
Agreeableness, Extroversion & Consciousness, Extroversion & Emotional Stability, Extroversion &
Imagination ,Agreeableness & Consciousness, Agreeableness & Emotional Stability, Agreeableness &
Imagination, Consciousness & Emotional Stability, Consciousness & Imagination, Emotional Stability &
Imagination and the analysis is done
32. Paired t-test R Code:
๏ # R Code for Paired T Test & Sample output:
๏ > library("duplr")
๏ > library(โpaireddata")
๏ > library(โduplr")
๏ t.test(data30$`Emotionional Stability`,data30$Imagination,
๏ + paired = TRUE, alternative = "two.sided")
๏ Paired t-test
๏ data: data30$`Emotionional Stability` and data30$Imagination
๏ t = -1.886, df = 29, p-value = 0.06935
๏ alternative hypothesis: true difference in means is not equalto
0
๏ 95 percent confidence interval:
๏ -5.3500924 0.2167591
๏ sample estimates:
๏ mean of the differences
๏ -2.566667
34. Conclusions
โข Extroversion & Agreeableness:
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the extroversion and agreeableness of the students .
โข Extroversion & Consciousness
Since p-value >0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Extroversion & Consciousness of the students .
โข Extroversion & Emotional Stability
Since p-value >0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in Extroversion & Emotional Stability of the students .
โข Extroversion & Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Extroversion & Imagination of the students .
โข Agreeableness & Consciousness
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Agreeableness & Consciousness of the students.
35. Conclusions
โข Agreeableness & Emotional Stability:
Since p-value >0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the agreeableness and emotional stability of the students .
โข Agreeableness & Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Agreeableness & Imagination of the students .
โข Consciousness & Emotional Stability
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the consciousness and emotional stability of the students.
โข Consciousness & Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Consciousness & Imagination of the students .
โข Emotional Stability & Imagination
Since p-value > 0.05 , we accept the null hypothesis and hence we conclude that there is no significant
difference in the Emotional Stability & Imagination of the students .
36. Limitations
โข Peer pressure, leading to not filling thesurvey
accurately
โข There was no compulsion to fill in the survey. As a
result it leads to lack of data , and hence results
could be misleading .
โข The employed psychometric test is of general
purpose and old ,newer industry-oriented
psychometric test could be employed.
37. Future Scope
The main scope of this project is to empower students and the
organizations ,which could be achieved in ways like:
โข Tracking group of students who lack in previously discussed
psychometric domains and hence the organizations can arrange
workshops or other activities to boost the skills of the students.
โข Assess the student's psychometric profile with the existing job
domains and hence counsel the student for a better career decision.
โข Promote studies and research in psychological statistics to provide
the student with experience and exposure for this emerging field.
Future Scope
38. Acknowledgement
Lastly, we would take this opportunity to thank Our Mentor
Prof. Kavya S and Prof. Dr. Santosh D for their humble
guidance and time to time involvement throughout the project
as well as for taking their time to teach us about all topics
covered in project without which It would be possible to
compile the project .
We would also like to thank Our Dean Preethi Maโam and
all our faculty professors for their suggestions and critical
assessment which helped us improve on our short comings in
our project and presentation.
We are highly grateful to All Students of NMIMS
Bengaluru to participated in survey and generously provided
us the data for our project