Category: Engineering

You have spent six weeks in the materials laboratory. There are cube crushing results at 7, 14, 21 and 28 days sitting in a notebook, a sieve analysis you ran three times, and a slump reading for every mix. Now you open Chapter Four to start writing, search for guidance, and every template you find opens with a bio-data table of respondents, a four-point Likert scale and a 2.50 decision criterion.

None of it applies to you. You did not administer a single questionnaire and you are not going to.

This is one of the loneliest points in a Nigerian engineering project, and it is entirely solvable. Chapter Four does the same job in every discipline — it presents what you found and interprets it — but in civil engineering the data are measurements, the comparison is against a standard rather than a mean criterion, and the tables look completely different.

What Chapter Four is actually for

Strip away the survey conventions and Chapter Four answers three questions in order:

1. What did the tests produce? — presentation
2. What do those numbers mean against a reference? — analysis
3. Why did they come out that way? — interpretation

Your chapter follows your objectives from Chapter One, in the same order, exactly as a survey-based project does. If objective one was to determine the physical properties of the material and objective two was to determine the compressive strength at varying replacement levels, then section 4.1 presents the physical properties and section 4.2 presents the strength results. That sequencing rule is the one thing engineering projects share with the standard structure described in the pillar on data presentation, analysis and interpretation — read it for the interpretation style, and ignore its Likert tables.

The four table types that carry an engineering Chapter Four

1. The material characterisation table

This comes first, before any performance result, because it establishes that your inputs were what you claimed. For a concrete study it holds specific gravity, bulk density, moisture content, fineness modulus and the aggregate grading. For a soil study it holds natural moisture content, specific gravity, Atterberg limits — liquid limit, plastic limit, plasticity index — and the grading from the sieve analysis, followed by the AASHTO or USCS classification that those values produce.

Each row should carry three columns: the property, your measured value, and the standard requirement or typical range. That third column is what turns a data dump into analysis, and students consistently leave it out.

2. The results-by-variable table

This is the heart of the chapter. Rows are your treatment levels — replacement percentages, curing ages, mix ratios, compaction efforts — and columns are the measured outputs.

For a partial replacement study, a table of average compressive strength in N/mm² with replacement level down the side and curing age across the top presents the entire experiment on a single page. Report the mean of your three cubes per age per mix, and state in the table caption that each value is the average of three specimens. Panels ask whether you tested one cube or three, and a caption answers it before the question arrives.

3. The comparison-against-standard table

Every measured value in a civil engineering project means something only relative to a specification. Your comparison column names the reference explicitly — the relevant British Standard for the test method, the Nigerian Industrial Standard for the material, or the Federal Ministry of Works General Specification for Roads and Bridges for pavement layer requirements.

So a CBR value of 12 per cent is not a finding on its own. “A soaked CBR of 12 per cent, which falls below the minimum specified for subbase material in the Nigerian General Specification, indicating that the soil requires stabilisation before use in that layer” is a finding, because it carries a decision.

4. The trend chart

One chart per relationship, no more. Strength against curing age with a line for each replacement level. Dry density against moisture content for a compaction curve, with the maximum dry density and optimum moisture content marked. Load against deflection for a structural test.

Axes labelled with units, a caption below in your department’s figure-numbering format, and referenced in the text as “Figure 4.3 shows…” before it appears. Charts that are never mentioned in the prose read as decoration.

Do you need statistics at all?

Often less than you fear, and occasionally more.

If your project measured properties without comparing treatments — characterising a soil, testing sandcrete blocks from three markets — descriptive presentation against standards is sufficient. Means and, ideally, standard deviations across your replicates are all a panel expects.

If your project varied a treatment across levels, an analysis of variance strengthens the chapter considerably. Testing five replacement percentages at 28 days, ANOVA tells you whether the differences between mixes are statistically significant rather than experimental scatter, and a post-hoc test tells you which specific pairs differ. Regression is appropriate where you are modelling one continuous variable against another, such as strength against curing age.

None of this requires expensive software. The tools that handle ANOVA and regression on a project-sized dataset are compared in the piece on SPSS, Excel, JASP and jamovi for a final year project, and two of the four are free.

The interpretation paragraph, in three moves

Presenting a table is not analysis. After every table, write a paragraph that makes three moves:

Move one — state what the table shows, in numbers. “Table 4.2 shows that compressive strength at 28 days decreased from 24.6 N/mm² at 0 per cent replacement to 18.3 N/mm² at 20 per cent replacement.”

Move two — compare it to the reference or the trend. “The 10 per cent replacement mix attained 22.1 N/mm², which exceeds the 20 N/mm² target strength for the design grade, while replacement levels above 15 per cent fell below it.”

Move three — explain why, with a citation. “The reduction at higher replacement levels is attributable to the lower calcium content of the pozzolanic material and its slower rate of strength development, consistent with the findings of [author] ([year]), who reported a similar decline beyond 15 per cent replacement.”

Three moves, three or four sentences, after every single table. Do that consistently and Chapter Four writes itself at roughly a page per table.

What belongs in Chapter Five instead

Recommendations do not go in Chapter Four. “Ten per cent replacement is recommended for non-structural applications” is a Chapter Five sentence. Chapter Four reports that the 10 per cent mix met the target strength; Chapter Five decides what should be done about it, alongside the summary, conclusion and contribution to knowledge set out in the guide to writing Chapter Five of a Nigerian final year project.

Keeping that boundary clean is one of the easiest marks in an engineering project, and mixing the two is one of the most common corrections.

The questions the panel will ask about your results

Four, reliably:

• “How many specimens per data point?” Three is the expectation for cube tests. If you tested one, say so and record it as a limitation rather than letting the panel discover it.
• “Which standard did you test to?” Name the test standard, not just the test. “Compressive strength was determined in accordance with BS 1881” is the answer.
• “Why did the strength drop at that level?” A mechanism, not a restatement. Chemistry, particle packing, water demand — something physical.
• “What would you have done differently?” Have one honest answer ready: more curing ages, a wider replacement range, duplicate soaked CBR samples.

The wider inventory of what Nigerian panels ask, across all six areas they probe, is in the guide to project defence questions in Nigeria. Computer science students face the same structural problem with a build project instead of a laboratory one, and the parallel is worth reading in the comparison of system development methodologies for a computer science project.

You have the results. The writing is the bottleneck.

This is the part nobody warns engineering students about. The laboratory work is finished, the numbers are correct, you know exactly what they mean — and there are still forty pages of formal departmental prose standing between you and submission, with a defence date that is not moving.

Writing up is not the same skill as running a CBR test, and it should not be what costs you a class of degree.

Tesify drafts your chapters from your own topic, objectives and results. It writes around the tables you already produced, keeps your terminology and reference list consistent across every chapter, and lets you rework a section after your supervisor’s corrections without rebuilding the chapter from the beginning. Your data stay yours, your findings stay yours, and the interpretation stays yours — what disappears is the blank page.

The free plan is the place to start. You can draft a chapter, see whether the output sounds like your department’s house style, and decide from there without paying anything.

Start your project on Tesify for free and get Chapter Four written around your test results this week.

Frequently asked questions