Statistics Practical (SPSS)
Statistics Practical (SPSS)
3 Practical assignment
Introduction
independent education committee based on a report prepared by the management department which
contained important statistics, such as: the students’ characteristics, performance on statistic
courses and student evaluation on said courses. These statistics were researched by a group of able
However, the research team of the department primarily conducted their study based on the
research questions provided by the management department. The first was a determination whether
there was a difference in courses 1.3 and 2.2 in statistic grades between students who only gained
credits for the first two years of the course, despite being half-way through the third year, and
students who accumulated more credits during the entire course. The following hypotheses were
formulated: Null Hypothesis: There were no differences between students who gained credits in the
first two years and the students who gained more credits. Alternative Hypothesis: There were
differences between students who gained credits in the first two years and the students who gained
more credits. The second question was whether there were differences in statistic grades, in course
2.2, between students who were satisfied with the programme quality to those who were
dissatisfied. The following hypotheses, for this question, were formed. Null Hypothesis: There
were no statistical differences between students who were satisfied and dissatisfied with the
programme quality. Alternative Hypothesis: There were statistical differences between students
Aside from testing the prior mentioned hypotheses, the relationship of the two courses
which formed the basics of Methodology and Statistics, the students’ main characteristics and the
Method
Practical assignment
Please type your answer within the borders of the TEXT BOXES below. You are not allowed to change the format, text outside the
boxes will not be graded.
Method
Participants
For this research, a random sample of n=210 students was drawn from a total of 5,000
students. The gender ratio, as charted in a pie graph, showed that the 56.67% were males, making it
the largest proportion, whereas 43.33% were females. The age of the participants, ranged from 22
to 29 years (Mage=25.5 years, SD=1.19). The age distribution showed, that the min. value was 22
years, the 25th percentile was 24 years and below, the median was 25 years, the 75th percentile was
26 years and above, and the max. value was 29 years. On a histogram, at which the age was charted
in, it was shown that the distribution is approximately normal; just skewing slightly to the right.
The programme ratio, as charted in a bar graph, showed that 47.1% of participants were enrolled in
Psychology (Dutch); making it the highest peak. The next highest peak, was 31.0% of participants
enrolled in Pedagogy. Lastly, the lowest peak was 21.9% of participants enrolled in Psychology
(International). The Living situation of the participants, as charted in a bar graph, showed that
47.6% live with peers/partners, making it the highest peak. The second highest peak was 33.8% of
participants living alone, then 16.7% of participants living with their parents, and finally 1.9% have
another form of living situation; making it the lowest peak on the graph.
OSIRIS is an online study information platform that holds records of the participants’ data;
those being their age, gender, credits, grades, programme and living situation. Through this
platform, the data was made available for students and administrators alike. All participants were
informed of the research and its purpose, and with their written and informed consent, they were
handed questionnaires with seven items. Each item had a scale from 1 through 10, at which the
participants would have to select based on their opinions. One of the items were: “Please indicate
the overall quality of the program you are following”, the data on programme satisfaction was
collected this way. All the available and questionnaire data were recorded and analyzed onto a
program known as Statistical Package for Social Sciences (SPSS) version 25. The statistical tests
on that data that were used, included: descriptive statistics (frequencies, descriptive, and crosstabs),
compare means (independent samples t-test), and correlation (bivariate, as we were comparing the
Results
A bivariate analysis has shown a correlation existing between the grades of 1.3 and 2.2. The
correlation was about r=.553, indicative of a moderate and linear correlation. The points on the
scatterplot do not cluster as much in relation to the line of best fit, but a positive linear pattern does
show. There are a few outliers that deviate from the line of best fit. As the overall pattern suggests,
those who scored either high or low in course 1.3 tended to score higher in course 2.2, but a few
points stray far below from the line of best fit; this indicates that a low percentage of the
participants do not necessarily score better in later courses. However, another outlier scored
considerably low in course 1.3, but scored higher in course 2.2; indicating that a low percentage of
Table 1
Retake 1 3 5 9
Compensate 52 73 32 157
Safe 12 23 9 44
Total 65 99 46 210
Based on the crosstabulation, the marginal probabilities of each progress are as follows:
About 4.3 % (P=.043) of the participants must do a retake, those who must compensate are about
74.8% (P=.748), and the participants who are safe make about 21.0% (P=.021). Upon further
calculation, the conditional probability of being safe given the participants’ programme is as
follows. The probability that a student is safe, given they study in Psychology (Dutch) is P=.233.
The probability of being safe, given the participant studies Psychology (International) is P=.196.
Once again, the probability of being safe, given that the participant studies Pedagogy is P=.184.
These results reveal that a student being safe is not independent from either taking Psychology
(Dutch) or Psychology (International), as the probability results do not equal to the probability of
the majority being safe: Psy (Dutch)=.233 ≠ Safe=.21; Psy (International)=.196 ≠ Safe=.21.
Table 2
Above is a table with some descriptive statistics on both grades in each of the courses. One
notable observation is the min. and max. value differing in 1.3 and 2.2. This exhibited how the
grades varied in each course, that being how people scored lower in course 2.2 compared to 1.3.
The distribution of grades 1.3 was computed by calculating the percentile and z-scores for
each student. From the results, after computing, it showed that the percentile scores did not fully
coincide with the given z-scores. For instance, students who had a percentile score of 99.29 had a
z-score 2.75735 to which, when rounded off, should be three standard deviations away from the
mean. This indicated that while the distribution was not perfect, it was still approximately a normal
distribution. Another notable observation showed that most of the participants scored a 4 to 6.
The standard error is the estimated standard deviation from a set of data. By calculating by
hand, the formula is as follows: SE x̅ =s/√n. In relation to grades 2.2, the SE=.0705 and when
computed through SPSS, the result is the same. The confidence interval (CI), by hand, is calculated
by first finding the margin of error: m=z*(σ/√n). Next, one must add and subtract the margin of
error with the mean: x̅±m. By hand calculation, with a 95% CI, the population mean is located
between (5.993, 6.464). With SPSS, the result is still the same.
An independent samples test was used to test the first hypotheses. In course 1.3, about 123
of the participants had less than 120 credits (Mcredits= 5.547, SD=1.077) and 87 had more than 120
credits (Mcredits= 5.467, SD=.943). The mean scores of course 1.3 showed that the variances were
not equal and that the differences were significant, t(199)=.570, p=.167. In course 2.2, about 123 of
the participants had less than 120 credits (Mcredits= 6.184, SD=1.760) and 87 had more than 120
credits (Mcredits= 6.293, SD=1.698). The mean scores of course 2.2 showed that the variances were
equal and that the differences were not significant, t(208)=-.446, p=.639.
The same test was used on testing the second hypotheses. About 125 of the participants, in
course 2.2, scored a grade below 5.5 (Mgrades= 6.192, SD=1.749). The other 85 participants scored a
grade above 5.5 and higher (Mgrades= 6.283, SD=1.713). The mean scores of satisfaction showed
that the variances were equal and that the differences were not significant, t(208)=-.371, p=.750.
Conclusion
The study’s main aim was to research any differences in statistical grades in both courses
1.3 and 2.2, in terms of credit accumulation and satisfaction in course 2.2. Based on the data and
statistical tests, course 1.3 proved the alternative hypothesis correct, thus rejecting the null
hypothesis. Meaning that there was a difference with those who had less than 120 credits and those
who had more. For course 2.2, in terms of accumulated credits, the null hypothesis was proven
true. Meaning that there were no differences in how much credits were gained in this course.
In terms of the statistic grades of course satisfaction, the differences between the students
who either gained a score either less or greater than 5.5, yielded no significant difference. In this
case, the null hypothesis cannot be rejected. This indicated that regardless of what grade the
students had gained, it had no effect over their opinion on the course itself.
To summarize, the research team was able to identify differences in the credits gained by
students, at east in course 1.3 and not in 2.2. For the satisfaction of the programme, in 2.2, they had
found no differences.