Project Report – “Method of Data Collection & Tools/Procedure”

Weightage : 20 Marks (as per typical academic rubric)

Below is a ready-to-use outline and content guide that you can drop into
the Methodology chapter of a project report ([Link] / [Link] / [Link] / [Link] / MBA, etc.).
Each sub-section is linked to the mark-allocation so you can see exactly what the
examiner expects and how to earn full points.

Section What the examiner looks for Marks

Introduction to the Brief context – why this method is appropriate

methodology for the research problem. 2

Research design Descriptive / exploratory / experimental /

(type & approach) correlational; qualitative, quantitative, or mixed. 2

Definition of target population, sampling frame,

Population & sampling technique, sample size calculation,
sampling justification. 4

Description of questionnaire,
interview-schedule, observation sheet, sensor,
Data-collection software, secondary-data sources, reliability &
instruments (tools) validity checks. 4

Data-collection Exact chronological steps, pilot testing, field

procedure work, mode of administration, ethical
(step-by-step) safeguards, handling non-response. 4

Coding, cleaning, software used (SPSS, R,

Data-processing & NVivo, Excel, Python, Tableau), statistical tests
analysis plan or thematic analysis. 2

Limitations of the Critical appraisal – what the method can’t

method capture, bias, reliability issues. 2

Total – 20
 Introduction to the Methodology (2 Marks)

“The methodology chapter explains how the research objectives will be achieved.”

 State the research problem in a single sentence.

 Explain why the chosen method (e.g., survey, experiment, case-study) is the
most suitable to answer the research questions/hypotheses.

 Cite one or two authoritative sources that support the choice

(e.g., Creswell 2018; Saunders et al., 2019).

Sample paragraph

“To assess the impact of digital-learning tools on undergraduate mathematics

performance, a cross-sectional survey design is adopted. This design enables the
collection of quantitative data from a large sample at a single point in time, which is
appropriate for testing the hypothesised relationship between tool usage frequency
(independent variable) and examination scores (dependent variable) (Creswell, 2018).”

Research Design (2 Marks)

Design When to use Key features

Descriptive You need prevalence / attitude Structured questionnaire,

Survey data from a large population. statistical analysis.

You manipulate an
independent variable and need Randomised control,
Experimental causal inference. pre-post tests.

Exploratory Little prior knowledge, aim to Semi-structured interviews,

Qualitative generate concepts. thematic analysis.

You need both breadth (quant) Sequential or concurrent

Mixed-Methods and depth (qual). design, integration matrix.

Write:

“The study employs a descriptive-survey design (type of quantitative research) …”

Population & Sampling (4 Marks)

Item What to include Example

Who you want to study

(e.g., all 2nd-year [Link]
Computer-Science
Target students at XYZ “All 2nd-year [Link] CS students
population University). (N ≈ 540) at XYZ University.”

List or source from which

the sample is drawn (e.g., “Enrollment database of the
Sampling university enrolment Department of Computer Science,
frame register). 2023-24.”

Probability (simple
random, stratified, cluster)
Sampling or non-probability “Stratified random sampling based
technique (convenience, purposive). on gender (Male : Female = 1.2 : 1).”

“Using Cochran’s formula with 95 %

confidence, 5 % margin of error,
Calculation (e.g., using p = 0.5, n ≈ 229. After adding 10 % for
Cochran’s formula) + non-response, the final sample =
Sample size justification. 252.”

Why this size/technique “Stratification ensures proportional

gives a representative gender representation, reducing
Justification outcome. sampling bias.”

Data-Collection Instruments (Tools) (4 Marks)

Tool Description Validation / Reliability

30 closed-ended items
(5-point Likert) covering Content validity checked by 3
usage frequency, subject-matter
perceived usefulness, experts; Cronbach’s
Questionnaire and self-e icacy. α = 0.86 from pilot test (n = 30).
 Tool Description Validation / Reliability

Google Forms / Platform compliance with

Online Survey Qualtrics – auto-scoring, GDPR; secure HTTPS
Platform export to CSV/Excel. connection.

If measuring digital-tool
usage, use Moodle System logs validated by
analytics API to extract cross-checking with manual
log data (login count, time-sheet for 5 participants
Sensor/Software time-on-task). (kappa = 0.92).

Institutional exam Obtained with permission from

scores (o icial the Registrar; anonymised by
Secondary Data transcripts). student ID.

Interview 5 open-ended questions

guide (optional, for probing barriers to Face validity confirmed
mixed-method) adoption. through expert review.

Tip: Include a sample questionnaire as an appendix (Appendix A). Show a screenshot

of the online form (Appendix B).

Data-Collection Procedure (Step-by-Step) (4 Marks)

1. Ethical clearance

 Submit proposal to Institutional Ethics Committee → obtain Ref.

No. EC-2023-017.

2. Pilot testing

 Administer questionnaire to 30 students → compute reliability; refine

ambiguous items.

3. Sampling & recruitment

 Generate stratified random list → email invitation with consent form &
survey link.

4. Informed consent
  Participants click “I Agree” before accessing the questionnaire (digital
consent recorded).

5. Administration

 Online mode: Participants complete the survey in their own environment;

average completion time ≈ 12 min.

 O line backup: For students lacking internet, paper copies are

distributed in the lab; later digitised.

6. Data capture

 Export responses from Qualtrics → CSV file → store on encrypted

university drive.

7. Monitoring response rate

 Send two reminder emails at 7-day intervals.

8. Handling non-response / missing data

 If < 80 % of items answered → discard; otherwise,

use mean-imputation for missing Likert items.

9. Data triangulation (if mixed-method)

 Conduct 10 semi-structured interviews (Zoom) → record → transcribe

using [Link].

10. Data security

 All files password-protected; access limited to principal investigator and

supervisor.

Flow-chart (optional) – Insert a simple diagram showing steps 1 → 10.

Data Processing & Analysis Plan (2 Marks)

Phase Action Software

Remove incomplete responses,

duplicate entries; code missing
Cleaning values. Excel + R (tidyverse)
 Phase Action Software

Frequencies, means, SD for each

Descriptive item; check normality
statistics (Kolmogorov-Smirnov). SPSS / R (psych)

Reliability
check Cronbach’s α for each scale. SPSS / R (ltm)

• Correlation (Pearson) between

tool-usage frequency and exam score.
Hypothesis • Multiple regression (control for SPSS
testing gender, prior GPA). (Analyze → Regression)

Qualitative (if Thematic coding, inter-coder

any) reliability (Cohen’s κ). NVivo / [Link]

Compare quantitative trends with Manual matrix in

Triangulation interview themes. Word/Excel

Reporting – Provide tables (Table 1: Demographics; Table 2: Reliability; Table 3:

Regression output) and figures (Figure 1: Histogram of usage scores; Figure 2: Scatter
plot with regression line).

Limitations of the Method (2 Marks)

 Sampling bias – reliance on voluntary participation may over-represent students

with higher digital-tool a inity.

 Self-reporting – Likert-scale responses subject to social desirability bias.

 Cross-sectional design – cannot infer causality; only associations.

 Technical constraints – Internet connectivity issues for a small subset of

respondents limited to paper mode.

How to mitigate – Mention remedial steps (e.g., anonymity to reduce desirability bias,
inclusion of objective log data to triangulate self-reports).

Putting It All Together – Sample Write-Up (≈ 800-900 words)

Below is a complete, ready-to-paste section that follows the rubric. Adapt the
institution-specific details, numbers, and tools as needed.

3. Methodology

3.1 Introduction

The present study aims to examine the relationship between the frequency of
digital-learning-tool usage and academic performance among undergraduate
Computer-Science students. A cross-sectional survey design has been selected
because it permits the collection of quantitative data from a sizable sample at a single
point in time, facilitating the testing of the hypothesised positive correlation
(Creswell, 2018).

3.2 Research Design

A descriptive-survey approach (quantitative) was adopted. This design enables the

measurement of variables (tool-usage frequency, perceived usefulness, self-e icacy,
and examination scores) using structured instruments and subsequent statistical
analysis.

3.3 Population, Sampling Frame & Sample

 Target population: All second-year [Link] Computer-Science students enrolled

in the 2023-24 academic year at XYZ University (N ≈ 540).

 Sampling frame: O icial enrollment list retrieved from the Department of

Computer Science’s registrar o ice.

 Sampling technique: Stratified random sampling based on gender to ensure

proportional representation (Male : Female ≈ 1.2 : 1).

 Sample size determination: Using Cochran’s formula with 95 % confidence

level, 5 % margin of error, and p = 0.5, the required size is n = 229. Adding a 10 %
contingency for non-response yields a final sample of 252 students.

The stratified approach reduces sampling error and enhances external validity.

3.4 Data-Collection Instruments

 Instrument Description Validation

30 items on a 5-point Likert

scale covering usage
frequency (e.g., “I use the Content validity verified by
digital learning platform at three faculty experts; pilot
least three times a week”), test (n = 30)
Online perceived usefulness, and produced Cronbach’s
questionnaire self-e icacy. α = 0.86.

Extracts log data (login

count, total time spent) for
each participant for the Cross-checked with manual
Moodle semester preceding data time-sheet for 5 participants
analytics API collection. (kappa = 0.92).

Final semester marks O icial records – no

Institutional obtained from the Registrar additional validation
exam scores (anonymised). required.

Five open-ended questions

probing perceived barriers;
used for a small
Interview mixed-methods supplement Face validity confirmed by
guide (optional) (n = 10). expert review.

The questionnaire and log data provide

complementary subjective and objective measures of tool usage.

3.5 Data-Collection Procedure

1. Ethical clearance – Approved by the Institutional Ethics Committee (Ref.

No. EC-2023-017).

2. Pilot testing – Conducted with 30 students; minor wording changes made based
on feedback.

3. Sampling & recruitment – A stratified random list of 252 students generated;

each received an email invitation containing an information sheet, consent form,
and a Qualtrics link.
 4. Informed consent – Participants clicked “I Agree” before accessing the survey;
digital consent was logged automatically.

5. Survey administration – Online mode; average completion time 12 minutes. For

participants without reliable internet (≈ 5 % of the sample), paper copies were
distributed in the computer lab and later digitised.

6. Data capture – Responses exported as CSV files and stored on the university’s
encrypted drive.

7. Reminder schedule – Two reminder emails sent at 7-day intervals to boost the
response rate (final response rate = 87 %).

8. Missing data handling – Items with < 80 % completion were excluded; remaining
missing Likert items were imputed with the respondent’s mean score for that
scale.

9. Interviews – Conducted via Zoom, recorded with permission, and transcribed

using [Link].

10. Data security – All files password-protected; access limited to the principal
investigator and supervisor.

A flow-chart of the above steps is presented in Figure 1 (appendix).

3.6 Data Processing & Analysis

 Cleaning – Duplicate and incomplete records removed; coding performed

in R (tidyverse).

 Descriptive analysis – Frequencies, means, standard deviations; normality

tested via Kolmogorov-Smirnov.

 Reliability – Cronbach’s α for each scale (α > 0.80).

 Inferential statistics – Pearson correlation between usage frequency (averaged

Likert score) and semester marks; followed by a multiple regression controlling
for gender and prior GPA. Analyses executed in SPSS (Version 28).

 Qualitative component – Thematic coding of interview transcripts in NVivo 13;

inter-coder reliability (Cohen’s κ = 0.88).

 Triangulation – Quantitative trends compared with emergent themes (e.g., “lack

of reliable internet” as a barrier).

Results are presented in Tables 1-3 and Figures 2-3 in Chapter 4.

3.7 Limitations
 1. Self-selection bias – Voluntary participation may over-represent highly
motivated students, potentially inflating the observed correlation.

2. Reliance on self-reported usage – Likert responses are susceptible to social

desirability bias, albeit mitigated by anonymity and corroboration with log data.

3. Cross-sectional nature – The design captures a snapshot and cannot establish

causality; longitudinal studies would be needed for causal inference.

4. Technical constraints – A minority of participants required paper-based

surveys, possibly introducing mode e ects.

Mitigation strategies (anonymity, mixed data sources, pilot testing) are discussed above.

4. Marking Checklist (for you)

Element How to show it in the report

Clear rationale for the chosen design 1-sentence “Why survey?” +

1 (2 pts) citation

2 Precise research design label (2 pts) “Cross-sectional survey design”

Population, sampling frame, technique, Bullet list + Cochran formula

3 size, justification (4 pts) figure

Detailed instrument description +

4 validation (4 pts) Table of tools + reliability α

Numbered list + flow-chart in

5 Step-by-step procedure (4 pts) appendix

6 Data cleaning & analysis plan (2 pts) Software names + test types

7 Critical limitations (2 pts) Bullet list + mitigation notes

If each bullet is present and well-written, you will secure the full 20 marks.

Quick Copy-Paste Template (≈ 300 words)

Methodology
The investigation employs a cross-sectional survey design to explore the link between
digital-learning-tool usage and academic performance among 2nd-year [Link]
Computer-Science students at XYZ University (N ≈ 540). A stratified random
sample (gender-proportionate) of 252 students was drawn using the department’s
enrollment register. Sample size was computed via Cochran’s formula (95 %
confidence, 5 % margin of error) and inflated by 10 % to accommodate non-response.
Data were collected through a 30-item Likert questionnaire (5-point scale) built in
Qualtrics, complemented by objective usage logs extracted from the Moodle analytics
API and o icial semester exam scores. Content validity was established by three
faculty reviewers; pilot testing (n = 30) yielded Cronbach’s α = 0.86.
The procedure comprised: (1) ethics approval (EC-2023-017); (2) pilot test and
instrument refinement; (3) email invitation with digital consent; (4) online survey
administration (average 12 min); (5) two reminder emails; (6) data export to CSV; (7)
cleaning (removal of incomplete cases, mean-imputation for ≤ 2 missing items per
scale); (8) storage on an encrypted university drive.
Analysis will be performed in SPSS: descriptive statistics, reliability testing, Pearson
correlation, and hierarchical multiple regression (controlling for gender and prior GPA).
Qualitative interview data (n = 10) will be thematically coded in NVivo. Limitations
include potential self-selection bias, reliance on self-report, and the cross-sectional
nature of the design. Mitigation strategies involve anonymity, triangulation with log data,
and rigorous pilot testing.