A computer science research proposal is marked on whether it shows a researchable problem, a method that can answer it and an honest awareness of its limits. It is not marked on how advanced the topic sounds. A modest, well-bounded question about phishing awareness, database performance or cloud adoption will usually score higher than an ambitious plan that cannot be carried out in one semester. This guide explains what markers look for in a task like the Applied Computer Science Research Project, which combines a proposal, a critical analysis of three papers and a short reflection.

If you would like to see the approach applied, our computer science research proposal sample is provided for reference only. You can read the full task on the Applied Computer Science Research Project brief.

What this kind of assessment is really asking

The brief usually asks you to choose a topic freely, identify a problem, challenge or opportunity, and plan a study without necessarily carrying it out. That last point matters. You are being assessed on design and thinking, not on building software or collecting a full data set. A typical split is a proposal of around 2,500 to 3,500 words worth the largest share of the marks, a paper analysis of roughly 1,000 to 1,200 words, and a reflection of 400 to 500 words.

Markers read for four things: a clear problem, a justified method, ethical care and critical use of sources. Descriptive writing that simply lists what authors said earns little. Comparing authors, spotting where they disagree and naming what is missing earns much more.

Choosing a topic that you can finish

Free choice is a trap for many students. A topic that is too broad, such as artificial intelligence in business, cannot produce a sharp question. Narrow it in three steps.

  1. Pick an area you can read about easily, such as cybersecurity awareness, machine learning for prediction, cloud adoption or usability.
  2. Choose a specific group or setting, for example university students or small companies in Malta.
  3. State one measurable question about it, such as whether a short training session changes behaviour.

Check feasibility before committing. Can you reach participants or find a public data set? Can you analyse the data with Excel or Python? Will ethics approval be straightforward? If the answer to any of these is no, change the topic now rather than after you have written the introduction.

How to split the proposal

Plan the word count before writing. The table below shows a workable distribution for a proposal of about 3,000 words. Adjust it to the limits in your own brief.

Section Suggested words What earns the marks
Title page and abstract 150 to 200 Problem, method and expected contribution in one paragraph
Introduction 300 to 350 Background narrowed to a specific problem
Aim, objectives, research questions 200 to 250 Questions that match the objectives one to one
Literature review 700 to 900 Comparison of authors and a clear gap
Methodology 500 to 600 Design, sample, instruments, justified choices
Ethics 250 to 300 Consent, privacy, confidentiality, data protection, responsible AI use
Data analysis plan, timeline, outcomes 350 to 450 Tool and technique matched to each question; realistic weeks

Write the aim as one sentence, then turn it into three or four objectives and two or three research questions. Every question should be answerable by a method you describe later. A common test is to draw a small grid with questions down the side and data sources across the top. If a question has no data, remove it.

Writing the literature review

The literature review is where most marks are lost. Group sources by idea, not by author. For each idea, say what the studies found, how their methods differ and why that might explain different results. A laboratory experiment, an online survey and a field study can reach different conclusions about the same behaviour, and pointing that out shows critical reading.

The brief usually asks for at least eight to ten credible sources, including journal articles and conference papers. Use databases and publisher pages, check each author list and year, and cite in the style the brief names, usually APA 7th edition. The citation and referencing help page explains how we approach APA formatting. Blogs and vendor reports can give context, but they are not a substitute for peer-reviewed work. Finish the section with a gap statement that leads straight into your aim.

Methodology, ethics and the analysis plan

Method

Match the method to the question. Surveys suit questions about attitudes and habits. Experiments suit questions about cause and effect. Interviews explain why people behave as they do. Case studies suit a single organisation. Secondary data suits questions that existing data sets can answer. Explain why you rejected the alternatives, because a justified choice scores higher than a fashionable one.

Ethics

Do not treat ethics as a paragraph of generic phrases. Say how participants will be recruited, what they will be told, how consent will be recorded and how they can withdraw. Describe how data will be pseudonymised, stored and deleted. In Malta, personal data are covered by the General Data Protection Regulation and the Data Protection Act, Chapter 586 of the Laws of Malta, so say how your plan respects both. If you use AI tools, state where and how you checked their output.

Analysis plan

Name the tool and the technique for each research question. Descriptive statistics and a chi-square test in Excel or Python suit many small studies. Interview data needs a stated approach such as thematic analysis. You do not need results, but you do need to show how you would interpret them.

Not sure whether your plan is realistic? You can submit your assignment brief and we will tell you what we can do with it.

Comparing the three approaches to the paper analysis

Part B asks you to analyse three papers on your topic, then compare them. For each paper, cover its purpose, method, data collection, findings, limitations and relevance to your study. The comparison is the part that carries the analysis.

Approach to the analysis What it looks like Likely result
Three separate summaries Each paper described in turn, no links between them Pass-level, descriptive
Summaries plus a short comparison One closing paragraph on similarities Mid-range
Comparison table plus synthesis Same headings for each paper, then discussion of contrasts in design and findings Strongest, because it shows critical reading

Choose papers that differ in method, for example one survey, one laboratory study and one field experiment. Differences give you something to compare. Read the methods section properly, not only the abstract, and note any limitation the authors state themselves before adding your own.

The reflection

Part C is short but easy to get wrong. Cover why you chose the topic, what made finding and judging sources difficult, what limits your proposal, what you would change with more time and how you checked credibility. Be specific. Saying that your sample is a volunteer group at one institution, so results cannot be generalised, is better than saying that more research is needed.

Common mistakes

  • Choosing a topic so broad that no single question can be asked.
  • Writing research questions that the chosen method cannot answer.
  • Listing sources in the literature review without comparing them.
  • Citing websites or vendor reports instead of journal and conference papers.
  • Describing ethics in general terms without consent, storage or deletion details.
  • Promising a software build or a data set that cannot be completed in the time available.
  • Giving a timeline with no weeks, or one that places writing up in the final days.
  • Summarising the three papers without comparing them.
  • Using AI tools without declaring them or without checking the output.
  • Leaving the reference list in mixed styles.

Mark-maximising checklist

  • The problem is specific and tied to a defined group or setting.
  • Aim, objectives and two or three research questions line up.
  • The literature review compares authors and ends with a gap.
  • At least eight to ten credible sources are used, including journal and conference papers.
  • The method is justified and the alternatives are acknowledged.
  • Ethics covers consent, privacy, confidentiality, data protection and responsible AI use.
  • Each research question has a named analysis tool and technique.
  • The timeline is realistic and shows weeks.
  • Part B compares the three papers with a table and a synthesis.
  • Part C names real limitations and how source credibility was checked.
  • Every in-text citation matches an APA 7th reference.

FAQs

Do I have to collect real data for the proposal?

No. A proposal plans the study. You describe the design, sample, instruments and analysis, but you are not normally expected to run the study or build software unless the brief says so.

How many sources do I need?

The brief for this type of task asks for at least eight to ten credible academic sources, including journal articles and conference publications. Quality matters more than the number, so each source should support a specific point.

Which topic is easiest to research?

One where you can reach participants or find open data, and where peer-reviewed studies already exist. Cybersecurity awareness, usability and technology in education often suit this because the literature is large and the methods are familiar.

Can I use Excel for the analysis plan?

Yes, if it fits the question. Excel handles descriptive statistics and simple tests, while Python or dedicated statistical software suits regression or larger data sets. State why your chosen tool is appropriate.

How do I write the ethics section if I am not collecting data yet?

Write it as a plan. Explain how you would obtain informed consent, protect privacy and confidentiality, store and delete data and seek approval from the relevant ethics committee before starting.

What does critical analysis of a paper mean in Part B?

It means judging the paper, not retelling it. Explain what it set out to do, how sound the method was, what the findings really show and where the limits lie, then compare it with the other two papers.

If you are short of time on a project like this, read about our dissertation writing service or the programming assignment help available for the analysis stage.