TFIT1003 Assignment: AI Tools in Education: Analyzing Student Experiences from a November 2023 Survey

Assignment Brief

Introduction

A survey on students’ experience of using AI tools was carried out in November 2023. The survey
was mainly among students in TU Dublin but some students from other universities were also
included. There are over 200 responses to the survey.

Outline of Brief

The aim of this assignment is to analyse the survey and make a report based on your analysis. The
report should be clear and easy to read giving some insights into the data collected from the survey.
Describe the demographics, usage etc.

Compare how people respond to two questions (see if there is a relationship between how they
answer one question vs how they answer the other). Use pivot table (or functions like countifs()) to
do the analysis in Excel. It is not required to do any statistical analysis. Simply show charts and
explain what you see.

As w ell as describing relationships between question, try to formulate a theory about the reasons
for the relationship (or lack of relationship, if none can be seen). This does not require references,
just use your intuition of what the reasons might be.

Specific Objectives

You will produce a report based on this survey. The report will have executive summary (v. short, 2
or 3 sentences), introduction (short), results and analysis (main part), conclusions (short summary of
main findings, possibly recommendations).
Analyse the data from the survey:

• Describe a few of the major features of the data: numbers, percentages, etc, of people in
various categories (not exhaustive, but those of interest, particularly where they relate to
some overall theme of the report). Utilise bar charts and other charts for presentation.

• Look at relationships between questions (variables). Use pivot tables or other methods to
compare responses between two questions (how the variables are related). You can choose
which questions to look at base some theory/intuition/hunch you may have about what
questions might be strongly related to other questions. Or you may just be interested to
explore the relationship of two questions You may also find that there is no visible
relationship – a less exciting, but valid, outcome of research work.

• Consider what the data is telling you. Explain what you see from the data and make sense of
this in simple terms. Construe some story (a theory) that might explain why the data is the
way it is. There is no need to back up theories with references.

The analysis in the report must include charts (pie, bar chart, stacked bar charts, etc) and/or tables
which show what you are describing. Charts must be clearly labelled: title, and as needed, axis labels
and legends. Use numbered captions, e.g. “fig 2.1”. Charts should be copied from Excel and pasted
as images into the report – this avoids issues with broken links between files after being sent by
email.

Most of the content will be in the results and analysis section, where you are describing the data and
features that you find. The conclusions section will summarise the main things you wish to highlight.
The executive summary is then a further condensing of what the main points of the report are.
Number sections and sub-sections of the report.

There is no need for references in this assignment. It is self-contained based solely on analysis of the
data.

Keep content short, brief explanations.

Don’t list numbers in the text. Numbers should be in tables or on charts. In the text just point out
key numbers and features from the tables and figures.
Max word count: 800

Include excel files of your analysis along with your report.

About the data file

The Excel file “AI Usage Survey data 2023 Nov (v3 Feb 2025).xlsx” contains the data. The first tab,
“QIndex” lists the questions. The tab “Data” contains the responses.

The tab “Data_2” is the same as “Data” except the responses to Q8 have been changed: on tab Data,
they are numbers (0 to 100); on tab Data_2, they are one of 5 levels “0.Not at all”, “1.Little”,
“2.Somewhat”, “3.Very”, “4.Extremely”. This will make analysis easier.

Q8, Q9 & Q10 have several sub-parts. In the QIndex tab, the main question is coloured blue. Each
sub-part has an identifier code, e.g. Q8.1 has code “Q8Undstnd”. Q8 main question is “How
important have AI tools been for the following” and Q8.1 has the text “Help me understand topic”.
In the Data tab, the rows 2 to 8 have the response options
The main table has headings which have question number and short codes. See the list of headings
in QIndex in the column “Q_id”. Below the main table, there are frequency counts given in row 250
to 256.

The corresponding labels for the frequencies are listed below them in rows 259 to 265. So for
example, for Q2Gender, there are 103 male and 121 female.

The tab “PivotExample” shows an example of a pivot table along with some example calculations.
A mean value for each of the gender groups is calculated in J33:R36. This could be used to compare
the two groups, seeing which group gave a higher value. Of course when the sample size is small or
the difference in value between the groups is small, it may not be possible to say there is a
significant difference.

Sometimes you may wish to combine groups of responses together. In J39:M43, the variable Q7Use
is condensed into two columns instead of five columns: “Low” and “Mid-High”, where “Low” adds
together “0.Never” and “1.Rarely” and “Mid-High” adds together “2.Monthly”, “3.Weekly” and
“4.Daily”. Calculations for Chi-square test are shown below also.