Quantitative Data analysis
Examples
By
Dr. Paul Netalisile Malunda
07/08/2025 Dr Paul Netalisile Malunda 1
Descriptive analysis of results
Teacher Disagree Non-committal Agree
Evaluation
HOD assesses the 369 (39.5%) 51 514 (55%)
way I teach (5.5%)
Setting targets with 391 (41.8%) 37 (3.9%) 507 (54.3%)
the HOD
Evaluations HOD 391 (41.8%) 51 (5.5%) 492 (52.7%)
based on the targets
set
Annual appraisal by 148 (15.8%) 57 (6.1%) 729 (78.1%)
the head teacher
Head teacher giving 277 (29.7%) 67 (7.1%) 590 (63.2%)
feedback on
appraisal
07/08/2025 Dr Paul Netalisile Malunda 2
Descriptive results
• The results indicate that slightly over 50 percent of the
teachers agreed with their subject heads at the
beginning of the academic term on the teaching and
learning targets and were appraised basing on these
targets.
• Although 78 percent of the teachers agreed that they
were annually appraised by the head teachers, a lower
percentage (63.2%) indicated that head teachers
discussed with them the results of the appraisals. This
implied that several teachers did not participate in
setting performance targets and some head teachers
did not give feedback on the appraisals undertaken.
07/08/2025 Dr Paul Netalisile Malunda 3
Example
correlation analysis of Pay and Output of ten
workers
Worker Pay Out put
1 18 15
2 16 16
3 16 14
4 14 12
5 13 12
6 12 13
7 10 8
8 7 6
9 5 3
10 3 2
07/08/2025 Dr Paul Netalisile Malunda 4
Using SPSS to determine the Pearson
correlation index
• Create an SPSS data file with Variable X for pay
and Y for out put
• Click analyze>correlate>Bi-variate
• Transfer X and Y to variable box
• Select Pearson correlation Coefficient. Select
two-tailed test of significance
• Click options to select for statistics needed on the
variables
• Press Continue>Ok
07/08/2025 Dr Paul Netalisile Malunda 5
Descriptive Statistics
out put
Mean Std deviation N
Pay 11.4000 5.0376 10
Out put 10.1000 5.0211 10
Mean for pay is 11.4 and the mean for output is 10.1. N stands for
number of respondent (workers used in the investigation) which is 10
07/08/2025 Dr Paul Netalisile Malunda 6
Correlations
significant at 0.01 level(2-tailed)
Pay Out put
Pay Pearson 1.000 0.973**
correlation
Sig(2-tailed) …… 0.000
N 10 10
Out put Pearson 0.973** 1.000
correlation
Sig(2-tailed) 0.000 ……
N 10 10
07/08/2025 Dr Paul Netalisile Malunda 7
Interpretation of results
• Pearson correlation coefficient of 0.973
means that there is a significant (**),
strong, positive relationship (since
0.973 is close to +1) between pay and
workers’ output. The calculated
significance value of 0.000 (less than
critical 0.01 sign value) means that this
relationship is significant.
07/08/2025 Dr Paul Netalisile Malunda 8
Coefficient of determination
• The coefficient of determination denoted
by symbol R2 gives the percentage of
variability in y that is accounted for by x.
The coefficient of determination is equal
to the coefficient of correlation squared.
07/08/2025 Dr Paul Netalisile Malunda 9
Regression analysis
Regression means working backwards.
We determine the value of Y from given
points. Hence the equation we get is
known as a regression function
Regression analysis provides estimates of
values of the dependent variable from
values of the independent variables
07/08/2025 Dr Paul Netalisile Malunda 10
Regression analysis
• The regression line is used to estimate
the values of the dependent variable
from those of the independent variable.
The equation of this line is known as the
regression equation
07/08/2025 Dr Paul Netalisile Malunda 11
Regression &correlation analysis
• The goal of a correlation analysis is to see
whether two measurement variables co
vary, and to quantify the strength of the
relationship between the variables,
whereas regression expresses the
relationship in the form of an equation
07/08/2025 Dr Paul Netalisile Malunda 12
Regression analysis
Quality of teaching Coefficient P-value
& learning
Teacher supervision 1.89 0.000
Teacher evaluation 1.54 0.000
Teacher professional 1.93 0.000
development
Sub-region 0.001 0.946
School status 0.15 0.481
Age 0.48 0.003
Gender -0.08 0.669
Marital status -0.05 0.702
Education level -0.45 0.005
07/08/2025 Dr Paul Netalisile Malunda 13
Analysis of results
• Results in the table indicate that, teacher
supervision, teacher evaluation, teacher
professional development, age, and education
level significantly explain quality of teaching
and learning.
07/08/2025 Dr Paul Netalisile Malunda 14
Multiple regression
• Results mean that, a unit increase in teacher
professional development would result into a
1.93 unit increase in quality of teaching and
learning. The results also indicate that with other
factors in the model held constant, a unit increase
in teacher supervision results into 1.89 unit
increase in quality of teaching and learning,
implying that quality of pedagogical practices
significantly increased with increased teacher
supervision.
07/08/2025 Dr Paul Netalisile Malunda 15
Regression analysis
• Similarly, a unit increase in teacher evaluation
with the other factors in the model held
constant results into a 1.53 unit increase in
the quality of teaching and learning.
07/08/2025 Dr Paul Netalisile Malunda 16
Example of t-test
Suppose children at the age of ten are
weighed and measured and we want to know
whether there is any difference in the average
heights of boys and girls at this age. We
perform an independent samples t-test. The
boys and girls are our two independent
samples- there is no connection between
them
07/08/2025 Dr Paul Netalisile Malunda 17
Example t-test
We start by stating our hypotheses
H0: there is no difference in heights between
boys and girls at 10 years
H1: there is some difference in the average
height of boys and girls at 10 years
07/08/2025 Dr Paul Netalisile Malunda 18
Performing the t-test using SPSS
Select
Analyze
Compare means
Independent samples t-test
And the t-test dialogue box appears
We move the height variable into the test
variable area
07/08/2025 Dr Paul Netalisile Malunda 19
Example
• The first table of the output gives some
summary statistics: the numbers in each
group, the mean, the standard deviation
and standard error of the mean of
heights
07/08/2025 Dr Paul Netalisile Malunda 20
Descriptive statistics
Sex N Mean Standard deviation Standard error mean
1:00 girl 647 53.287 2.5916 0.1019
2.00 boy 648 53.644 2.5264 0.0992
07/08/2025 Dr Paul Netalisile Malunda 21
Independent samples test
Leven’s test for T-test for equality of means
equality of
variances
F Sig t df Sig(2- Mean difference
tailed)
Equal 0.547 0.460 -2.512 1293 0.012 -0.3573
variances
assumed
Equal -2.512 1292.05 0.012 -0.3573
variances
not
assumed
07/08/2025 Dr Paul Netalisile Malunda 22
Independent samples t- test
The mean difference of heights is 0.3573. we
are testing whether this difference is
significantly different from zero. Looking at the
significance level of 0.012 which is less than
0.05, we conclude that there is evidence at 5%
level, to suggest that there is a difference
between the heights of boys and girls at age
ten. So we reject the null hypothesis and take
the alternative hypothesis
07/08/2025 Dr Paul Netalisile Malunda 23
Example
Suppose children at the age of ten are
weighed and measured and we want to know
whether there is any difference in the average
heights of boys and girls at this age. We
perform an independent samples t-test. The
boys and girls are our two independent
samples- there is no connection between
them
07/08/2025 Dr Paul Netalisile Malunda 24
Example
We start by stating our hypotheses
H0: there is no difference in heights between
boys and girls at 10 years
H1: there is some difference in the average
height of boys and girls at 10 years
07/08/2025 Dr Paul Netalisile Malunda 25
Performing the t-test using SPSS
Select
Analyze
Compare means
Independent samples t-test
And the t-test dialogue box appears
We move the height variable into the test
variable area
07/08/2025 Dr Paul Netalisile Malunda 26
Example
• The first table of the output gives some
summary statistics: the numbers in each
group, the mean, the standard deviation
and standard error of the mean of
heights
07/08/2025 Dr Paul Netalisile Malunda 27
Descriptive statistics
Sex N Mean Standard deviation Standard error mean
1:00 girl 647 53.287 2.5916 0.1019
2.00 boy 648 53.644 2.5264 0.0992
07/08/2025 Dr Paul Netalisile Malunda 28
Independent samples test
Leven’s test for T-test for equality of means
equality of
variances
F Sig t df Sig(2- Mean difference
tailed)
Equal 0.547 0.460 -2.512 1293 0.012 -0.3573
variances
assumed
Equal -2.512 1292.05 0.012 -0.3573
variances
not
assumed
07/08/2025 Dr Paul Netalisile Malunda 29
Independent samples t- test
The mean difference of heights is 0.3573. we
are testing whether this difference is
significantly different from zero. Looking at the
significance level of 0.012 which is less than
0.05, we conclude that there is evidence at 5%
level, to suggest that there is a difference
between the heights of boys and girls at age
ten. So we reject the null hypothesis and take
the alternative hypothesis
07/08/2025 Dr Paul Netalisile Malunda 30
Example- Anova
Suppose data has been collected on the
productivity (‘000 of Kgs) of farmers from
three educational backgrounds as shown in
the table below
Primary Secondary University
graduates graduates graduates
10 9 17
8 8 15
12 14 12
5 13 14
11 16
07/08/2025 Dr Paul Netalisile Malunda 31
Using SPSS to verify the hypothesis
1. click on Analyze>compare means>One Way
ANOVA
2. Transfer productivity to the dependent list
3. Transfer educational background to the factor
4. Click on options to select Descriptive and
other information required
5. Click continue >Ok to obtain the tables below
07/08/2025 Dr Paul Netalisile Malunda 32
Descriptive statistics for productivity by
educational background
n mean Std Std error
deviation
Primary 4 8.7500 2.9861 1.4930
graduates
Secondary 5 11.000 2.5495 1.1402
graduates
University 5 14.8000 1.9235 0.8602
graduates
Total 11.7143 3.4290 0.9164
07/08/2025 Dr Paul Netalisile Malunda 33
ANOVA table verifying significance of
productivity for educational background
Sum of df Mean F Sig
Squares square
Between 85.307 2 42.654 6.946 0.011
groups
Within 67.550 11 6.141
groups
Total 152.857 13
07/08/2025 Dr Paul Netalisile Malunda 34
Interpretation of results- Anova
• We compare the calculated F statistic (6.946)
with the F critical value (3.93) with 2 degrees
of freedom at 5% level (got from statistical
tables). The calculated value is greater than
the critical value, meaning that productivity of
farmers depend on educational back ground.
• Alternatively we could use the calculated
significance value (0.011) which is less than
the 0.05 leading to the same conclusion
07/08/2025 Dr Paul Netalisile Malunda 35
SPSS: Chi square goodness of fit
Example
Suppose there is a claim that students who go to
night clubs perform poorly in class. Data is
collected from 100 student and below are the
findings in the table
Poor performance
Yes No
Go to night Yes 38 15
clubs No 7 40
07/08/2025 Dr Paul Netalisile Malunda 36
Procedure
• Click analyze>Nonparametric tests>Chi-square
• Transfer poor performance to Test Variable
list. Select All categories equal for expected
values
• Click options for descriptive statistics
• Click continuous > ok to obtain the results
below
07/08/2025 Dr Paul Netalisile Malunda 37
Goodness of fit chi-square out put
Poor performance (yes/no)
Observed N Expected N Residual
No 55 50.0 5.0
Yes 45 50.0 -5.0
Total 100
55 of the students performed well, 50 were expected to
perform well, and the difference between the observed
and expected values is 5.0
Also 45 of the students did not perform well, 50 were
Expected not to perform well, giving a Residual of -5.0
07/08/2025 Dr Paul Netalisile Malunda 38
Goodness of fit chi square out put
Students who go to night clubs (yes/no)
Observed N Expected N Residual
No 47 50.0 -3.0
Yes 53 50.0 3.0
Total 100
In our sample 47 of the student never went to night clubs,
50 were expected to go to night clubs, and the difference
between the observed and expected values is -3.0
07/08/2025 Dr Paul Netalisile Malunda 39
Test statistics
Poor Go to night club
performance
Chi- 1.00 0.360
square
df 1 1
Asymp.sig 0.317 0.549
07/08/2025 Dr Paul Netalisile Malunda 40
Interpretation
(For the performance variable)
• The chi-square obtained is 1.00, with1 degree
of freedom and significance level of 0.317,
which falls well above the 0.05 alpha level.
This means that the difference between the
observed and expected values is not
significant. We can therefore conclude that
the number of students who perform poorly
doesn’t significantly differ from those who
perform well
07/08/2025 Dr Paul Netalisile Malunda 41
Interpretation
(for go to night club variable)
• The chi-square obtained is 0.360, with 1
degree of freedom and significance level of
0.549, which is above 0.05 alpha level. The
difference between the observed and
expected values is not significant. We
conclude that the number of students who
visit night clubs and those who do not does
not significantly differ. It is a 50%/50%
relationship
07/08/2025 Dr Paul Netalisile Malunda 42
