For information and reference only – not to be submitted as your own work.

This extract shows how an Applied Computer Science Research Project for the module Scientific Reading and Writing for Computer Science can be organised across its three parts: a research proposal, a critical analysis of three papers and a short reflection. The topic chosen here, phishing awareness among university students in Malta, is only one of the free choices the brief allows, and every figure marked illustrative is invented for demonstration. The full task is on the Applied Computer Science Research Project question page. If you want the reasoning behind the structure first, read our guide to writing a computer science research proposal.

Planning table

Part What the answer must do Weight Extract shown here
A. Research proposal Define a problem, aim, questions, literature, method, ethics, analysis plan, timeline and outcomes 70% Key sections in condensed form
B. Paper analysis Compare three papers critically, not summarise them 20% Comparison table and short synthesis
C. Reflection Explain topic choice, literature challenges, limitations and source credibility 10% Opening extract

Part A: Research proposal (extract)

Title

Phishing awareness and susceptibility among university students in Malta: a mixed-methods study of simulated phishing and training.

Abstract

Phishing remains a leading route into personal accounts and institutional systems, and university students use email, messaging and learning platforms every day. This proposal examines how well students recognise phishing messages, which personal and behavioural factors are linked to falling for them, and whether a short, attention-focused training session changes behaviour. A mixed-methods design is proposed: a questionnaire, a consented simulated-phishing exercise with a training comparison, and a small set of follow-up interviews. Quantitative data will be analysed in Python and Excel, and interview data will be coded thematically. The expected contribution is evidence, from a Maltese student context, that can inform proportionate awareness training.

1. Introduction

Phishing uses deceptive messages to persuade people to disclose credentials or take harmful actions. Technical filters stop many attacks, but those that arrive in an inbox still depend on a human decision. Evidence on who makes the wrong decision is mixed. A large online role-play study reported that people aged 18 to 25 were more susceptible than older groups, and that women were more susceptible than men (Sheng et al., 2010), while a workplace study found that messages carrying authority cues were more likely to be clicked (Williams et al., 2018). Students sit in the age band that appears most at risk, yet sources reviewed for this proposal do not include a study centred on students in Malta. That is the problem addressed here.

2. Aim and objectives

Aim: to establish how susceptible university students in Malta are to phishing and whether brief training improves their behaviour.

  • To measure students’ ability to distinguish phishing from legitimate messages.
  • To identify behavioural and demographic factors associated with clicking in a simulated exercise.
  • To compare two short training formats on behaviour after ten days.
  • To explore, through interviews, why students decide as they do.

3. Research questions

  1. RQ1: How accurately do students identify phishing messages, and which factors are associated with errors?
  2. RQ2: Does attention-focused training reduce clicks on simulated phishing messages more than rule-based training?
  3. RQ3: What reasons do students give for trusting or distrusting a suspicious message?

4. Literature review (extract)

Early work established that the user, not only the software, is the weak point. Kumaraguru et al. (2007) tested an embedded training email system in laboratory experiments and found that it performed better than standard security notices about phishing. Sheng et al. (2010) used a role-play survey with 1,001 online respondents and reported that educational materials reduced the tendency to enter information on phishing pages by 40%, although some materials also made people slightly less willing to click legitimate links. The two studies agree that training helps, but differ in setting and in what they measure: one observes behaviour in a laboratory, the other relies on role-play, where nothing real is at stake.

Later work asks why people still fail. Alsharnouby et al. (2015) tracked the eyes of 21 adults judging 24 websites and found that participants correctly identified only 53% of phishing sites, spending about 85% of their gaze on page content rather than on browser indicators. Vishwanath et al. (2011) took a theoretical route, testing individual differences in vulnerability within an integrated information-processing model. Jensen et al. (2017) moved from cues to attention, reporting that mindfulness-based training helped 355 participants avoid a simulated attack better than rule-based training.

Two gaps follow. First, the rule-based versus attention-based comparison has been run at a single US university, so transfer to other student populations is untested. Second, studies that measure behaviour rarely pair it with students’ own explanations. This proposal addresses both by adding interviews and a Maltese sample.

5. Research methodology

Design. An explanatory sequential mixed-methods design. The quantitative phase comes first, and the interviews then explain patterns in it. The proposal tests two interventions and does not require building any software.

Participants and sampling. Undergraduate students at a Maltese higher-education institution, recruited by open invitation. A power calculation for a three-group comparison (control, rule-based, attention-based) with a medium effect (f = 0.25), alpha .05 and power .80 gives roughly 159 participants in total (Faul et al., 2007). The figure is illustrative and will be recomputed in G*Power once the final design is agreed. About 10 to 12 volunteers will be interviewed.

Instruments. (1) A short questionnaire on demographics, email habits and confidence. (2) A message-classification task of 12 screenshots, half phishing, half legitimate. (3) A simulated phishing exercise using a harmless landing page that records only whether a link was opened, never any typed input. (4) A semi-structured interview guide.

Procedure. Participants complete the questionnaire and classification task, receive one of three conditions, and ten days later receive the simulated message. Interviews follow within two weeks.

Condition Content (illustrative) Length
Control General institutional email notice 2 minutes
Rule-based Checklist of cues: sender, link, urgency, attachments 10 minutes
Attention-based Pause, notice the context, withhold judgement, then check 10 minutes

Our programming assignment help page is a useful starting point if your own proposal needs a small Python script for the analysis stage.

6. Ethical considerations

The study will seek approval from the institution’s research ethics committee before any contact with participants. Participation is voluntary and based on informed consent, with a plain-language information sheet that explains the purpose, what is recorded and the right to withdraw. Because a simulated phishing message involves a degree of deception, participants will be told at recruitment that they will receive realistic messages during the study, without being told which or when, and will be debriefed in full afterwards. Responses will be pseudonymised, stored on an encrypted institutional drive and reported only in aggregate.

Personal data are governed by the General Data Protection Regulation (Regulation (EU) 2016/679) as further implemented in Malta by the Data Protection Act, Chapter 586 of the Laws of Malta, with the Information and Data Protection Commissioner as the supervisory authority (Data Protection Act, 2018; European Parliament and Council, 2016). Only data needed to answer the questions will be collected, a retention period will be stated, and no landing-page entries will be stored. Responsible use of AI is also covered: any AI tool used for drafting or coding will be declared and its output checked by the researcher.

7. Data analysis plan

Question Data Technique Tool
RQ1 Classification scores, questionnaire Descriptive statistics; regression on error rate Python (pandas, statsmodels)
RQ2 Click or no click by condition Chi-square test; logistic regression with confidence as covariate Python, Excel for checks
RQ3 Interview transcripts Six-phase thematic analysis (Braun & Clarke, 2006) Spreadsheet coding grid

Interview findings will be set beside the statistical results in a joint display, so that a pattern such as higher clicking among confident students can be explained by what those students said.

8. Project timeline (illustrative)

Weeks Activity
1–3 Literature review, ethics application, instrument design
4 Pilot with five volunteers and revisions
5–8 Recruitment, questionnaire, training sessions
9–10 Simulated message, debriefing
11–12 Interviews and transcription
13–15 Analysis and integration
16 Writing up and submission

9. Expected outcomes

The study should deliver a descriptive picture of student susceptibility in Malta, an estimate of the difference between the two training formats, and a thematic account of student reasoning. A practical product is a one-page set of recommendations for the institution’s IT service. Findings will be presented as indicative, given a volunteer sample.

Part B: Critical analysis of three papers (extract)

Sheng et al. (2010) Alsharnouby et al. (2015) Jensen et al. (2017)
Venue CHI conference International Journal of Human-Computer Studies Journal of Management Information Systems
Purpose Who is susceptible, and do interventions work How users judge websites and where they look Whether mindfulness training beats rule-based training
Method and data Online role-play survey; 1,001 respondents Laboratory task with eye tracker, interviews; 21 adults, 24 websites Field experiment with simulated attack ten days after training; 355 participants
Main finding Ages 18 to 25 and women more susceptible; materials cut information entry by 40% 53% of phishing sites identified; about 6% of gaze on browser indicators Mindfulness training avoided attacks better than rule-based training
Limitation (my reading) Role-play carries no real consequences Participants knew they were being tested, so results are an upper bound; small sample Single university; one attack at one point in time
Relevance to my study Justifies a student focus and a training comparison Supports a classification task and interviews Direct model for the two training conditions

The three papers agree that human judgement is the weak point and that training can help. They differ in evidence: the survey measures stated choices, the laboratory study measures attention, and the field experiment measures behaviour toward a real message. The findings also point in different directions on mechanism. Alsharnouby et al. suggest people look in the wrong places, whereas Jensen et al. suggest that the way people allocate attention matters more than knowing the rules. Only the field experiment tests behaviour outside a prompted setting, which is why my design borrows from it while adding interviews to explain the choices that numbers cannot.

Part C: Research reflection (extract)

I chose phishing because every student receives suspicious messages and few have been shown how they decide. The main challenge in the literature was comparability: studies used different measures, such as role-play choices, gaze time and clicks, so I could not simply set their percentages side by side. I handled this by comparing designs before results. The proposal is limited by a volunteer sample, a short follow-up of ten days and the fact that students who know they are in a security study may behave more carefully. With more time I would add a second measurement after three months. To judge credibility I preferred peer-reviewed journal and conference papers, checked each DOI, and read the methods rather than relying on abstracts alone.

Students who want feedback on a structure like this before submission can send us their brief. For wider support with research writing, see our dissertation writing service, the citation and referencing help page and related samples on enterprise information systems and research methods.

References

Alsharnouby, M., Alaca, F., & Chiasson, S. (2015). Why phishing still works: User strategies for combating phishing attacks. International Journal of Human-Computer Studies, 82, 69–82. https://doi.org/10.1016/j.ijhcs.2015.05.005

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa

Data Protection Act, Cap. 586, Laws of Malta (2018). https://idpc.org.mt/our-office/legislation/

European Parliament and Council. (2016). Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data (General Data Protection Regulation). Official Journal of the European Union, L 119.

Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146

Jensen, M. L., Dinger, M., Wright, R. T., & Thatcher, J. B. (2017). Training to mitigate phishing attacks using mindfulness techniques. Journal of Management Information Systems, 34(2), 597–626. https://doi.org/10.1080/07421222.2017.1334499

Kumaraguru, P., Rhee, Y., Acquisti, A., Cranor, L. F., Hong, J., & Nunge, E. (2007). Protecting people from phishing: The design and evaluation of an embedded training email system. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 905–914). ACM. https://doi.org/10.1145/1240624.1240760

Sheng, S., Holbrook, M., Kumaraguru, P., Cranor, L. F., & Downs, J. (2010). Who falls for phish? A demographic analysis of phishing susceptibility and effectiveness of interventions. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 373–382). ACM. https://doi.org/10.1145/1753326.1753383

Vishwanath, A., Herath, T., Chen, R., Wang, J., & Rao, H. R. (2011). Why do people get phished? Testing individual differences in vulnerability within an integrated, information processing model. Decision Support Systems, 51(3), 576–586. https://doi.org/10.1016/j.dss.2011.03.002

Williams, E. J., Hinds, J., & Joinson, A. N. (2018). Exploring susceptibility to phishing in the workplace. International Journal of Human-Computer Studies, 120, 1–13. https://doi.org/10.1016/j.ijhcs.2018.06.004