Skip to content
+44 7386783749

Data Skills Coursework: From Raw Data to Clear Findings

Spreadsheet of ride times with one outlier highlighted

Published 10 Oct 2026 · Updated 9 Oct 2026

Data skills coursework asks you to take a messy dataset, prepare it, analyse it and explain the result to a reader who does not code. This guide gives a working order, a small calculated example and a spreadsheet checklist. You do the analysis; a mentor can comment on your plan.

Get feedback on your analysis plan

What a data skills module marks

Module codes such as MOD008862 differ between universities, so read your own brief and marking grid first. Most data skills modules reward the same four things.

  • Preparation: the data is checked, cleaned and the changes are recorded
  • Method: the summary or test fits the type of data and the question
  • Presentation: tables and charts are labelled and make one point each
  • Judgement: the write-up says what the numbers mean and where they fall short

Check how the marks are split between the technical file and the written part. A perfect spreadsheet with a vague explanation rarely scores well.

Data skills coursework: a working order

Work through these steps in order and keep notes as you go. Each one becomes a paragraph in your method section.

  • Turn the brief into two or three questions the data can answer
  • Save an untouched copy of the original file and work on a second copy
  • Look at the columns: meaning, type, units and where the data came from
  • Clean it and write down every change in a log
  • Summarise it, then draw the charts that answer your questions
  • Run a test only if the brief asks for one and the data allows it
  • Write up, then check each claim against a table or chart

Data analysis coursework: a worked example

Here is an invented bike-hire dataset. Eleven ride times, in minutes, were recorded at one docking station. Ten look ordinary and one reads 95.

MeasureAll 11 ridesWithout the 95
Mean17.710.0
Median109.5
Standard deviation25.83.0
Range6 to 956 to 15

One value moved the mean by almost eight minutes and made the spread look huge. The median barely moved. Before deleting anything, ask why it is there.

  • A bike left docked in the wrong place suggests an error, so flag it and say so
  • A genuine long ride is real data, so keep it and report the median
  • Show both sets of figures when you cannot tell which case you have

Markers reward the reasoning, not the choice. Our guide to ordinal and interval data explains how the type of data limits which averages make sense.

Data skills assignment: keep a cleaning log

A cleaning log shows a marker exactly what you changed. This invented extract matches the bike example.

IssueActionRows affectedReason
Duplicate ride IDsRemoved the second copy3Same ID, same times
'Stn 4' and 'Station 4'Standardised to Station 412One location, two spellings
Dates stored as textConverted to date formatAllNeeded for monthly totals
Ride of 95 minutesFlagged, kept in a second column1Cannot confirm an error

Put the full log in an appendix and summarise it in two sentences in your method.

Excel coursework help: tasks and checks

The table lists common spreadsheet tasks and the check that stops a silent mistake.

TaskToolCheck before you move on
Remove repeatsData > Remove DuplicatesWork on a copy and compare row counts
Look up a valueXLOOKUP, or VLOOKUP in older versionsTest three rows by hand
Group and totalPivotTableDo the totals match a simple SUM of the column?
Show a trendLine chartAxis labelled, units stated, title says the finding
Prepare the data for a PivotTable

Microsoft advises one header row with a unique label per column, no blank rows or columns and one data type per column. Turn the range into an Excel table so new rows are included when you refresh.

XLOOKUP and older Excel versions

XLOOKUP is not available in Excel 2016 or 2019. If your university computers run an older version, use VLOOKUP or INDEX with MATCH. XLOOKUP matches exactly by default and returns #N/A when nothing is found, so decide what that should show.

Check a PivotTable against a plain formula

Pick one group and calculate its total with SUMIFS or COUNTIFS. If the figures differ, there is a hidden blank, a spelling variant or a filter you forgot.

Choosing charts that make one point

  • Compare categories with a bar chart that starts at zero
  • Show change over time with a line chart
  • Show spread with a histogram or box plot
  • Show a link between two numeric variables with a scatter plot

Give every chart a title that states the finding, such as 'Weekend rides last longer than weekday rides'. Use colours that stay readable in greyscale. UK government analysts publish chart and accessibility guidance; see the sources below.

Personal data and ethics

If your dataset describes people, UK data protection law applies. The ICO lists seven principles. They cover lawfulness, fairness and transparency, purpose limitation, data minimisation, accuracy, storage limitation, integrity and confidentiality, and accountability. In practice, use a public or synthetic dataset where you can, drop columns you do not need and store files securely. Removing names may not be enough, because the ICO notes that effective anonymisation depends on the technique. Read ethical considerations in research if you collect your own data.

Writing up your findings

Open each results paragraph with the finding, then point to the table or chart that shows it. Keep methods in the past tense and limits in plain words.

A sentence frame for results

'Median ride time was 10 minutes (n = 11). The mean was higher, at 17.7, because one ride lasted 95 minutes (Table 1).' It gives the figure, the sample size and the reason for the gap.

A sentence frame for limits

'The data covers one station for one week, so the pattern may not hold elsewhere.' Name the gap, not a long apology.

Where the written part sits

Use the structure your brief sets. If it leaves the layout open, how to write a report gives a clear order, and writing the results section helps with presenting figures.

Common mistakes

  • Reporting a mean for skewed data without the median
  • Deleting outliers without saying so
  • Charts with no units, or a truncated axis on a bar chart
  • Pasting spreadsheet output with no sentence explaining it
  • Claiming one thing caused another because they moved together

Need feedback on your analysis plan?

A mentor can comment on your questions, method and draft write-up, and you complete and submit the work yourself. See how it works and our academic integrity policy. For testing group differences, one-way ANOVA explained is a good next read.

Related guides

Sources

Frequently asked questions

What does a data skills coursework involve?

Usually a dataset, a set of questions and a report. You clean the data, summarise it, build charts, interpret them and explain your choices. Your brief sets the exact tasks.

How do I analyse data for coursework?

Start with your questions, check and clean the data, then summarise it with measures that suit each variable. Add a test only if the brief asks. Explain every figure in a sentence.

Can I get help with Excel data analysis?

You can get feedback on your plan, formulas and charts, and explanations of the method. You still build and submit your own workbook.

Should I remove outliers?

Not automatically. Find out whether the value is an error or a genuine result, then explain your choice. Showing results with and without it is often the safest route.

Can I use real personal data?

Check your university's ethics rules first. A public or synthetic dataset avoids most problems. If you must use personal data, keep only what you need and store it securely.

Does the module code matter?

Codes such as MOD008862 belong to a particular university or centre. Treat this page as general study help and follow your own brief and marking criteria.