Sarah’s teaching methods suit different learning styles. She uses inclusive language and encourages students to consider which groups are included and which excluded from datasets. Sarah adopts a decolonial approach to data from developing countries. To encourage students to participate, she awards 10% of the grade for engagement.

Teaching area

Lectures

Course

Introduction to Quantitative Research: core Level 7 introductory statistics and quantitative research methods module taken by students on three MSc degrees: Science and Technology Policy, Strategic Innovation Management and Sustainable Development. Around 50 mostly mature international students in their 30s and 40s, with work experience.

Description

The module has been running for several years. It’s my second year teaching it at Sussex, and this year I ran it with doctoral tutor Indrajit Adhikari. I have adjusted it to suit different learning styles and use these inclusive principles in my teaching practice.

• I ask students at the start of the year if they need any adaptations to the slides, lecturing style or recordings.
• We provide the materials a few days in advance of the teaching session.
• We give students a set of independent practice materials with solutions, so they can go through the examples again after the class.
• All the lectures and seminars are recorded and made available to all the students.
• I run a discussion board where I ask students to tell me about themselves, such as: ‘What are you reading right now? What are you going to do with your degree?’ When we know their interests, we try to find relevant examples that would engage them.
• I sometimes ask students how things work in their country, which can be inclusive. But I recognise that students might not want to be the spokesperson for their homeland. Some students may be scared that they would be reported to the authorities if they criticised their government in public.

Interrogating the data

When working with quantitative data, it’s important to think about the categories of people who are included, how they are referred to, and which categories may be left out. At the start of term, I explain that we frequently use health data. Demonstration public health datasets are openly available because the research is state funded. But there are issues with health data. It codes according to biological sex, for example. I try to update my use of language to be more inclusive, using terms such as ‘people with a uterus’ or ‘people who can get pregnant’. I point out that some health behaviours and preferences can have social stigma associated and are therefore subject to Social Desirability Bias (when respondents give answers that they think will make them look good, rather than the truth).

The appropriateness of certain terms is an evolving social norm. An example from the textbook considers how one party’s accessing of pornographic material influenced the longevity of their romantic relationship with another party. It is a well-worked example of the statistical tests involved, but I didn’t want to say ‘porn’ ten times in the lecture nor project tables with the variable label ‘porn’ on the screen. I used it last year without incident, but my thinking and also the prevailing social attitudes have changed, so this year I dropped the example.

I include data from developing countries, such as sustainable development data, but I make sure to adopt a decolonial perspective. I ask questions such as, ‘Was it collected by people in that part of the world? Does it serve their purposes? What can it teach us? Does anything in this dataset conflict with what a comparable dataset collected from ‘WEIRD’ countries would demonstrate?’.

I don’t always have time to rebuild the demonstration dataset, but if I can’t change the material, I can at least change how I talk about it. I remind the students to think about the source of the data, why it was collected and what language is used.

Sampling

We have a whole week when we discuss why sampling is important. We look at how people ‘lie’ with statistics. It could be an act of omission, a simple case of who is asked and who is left out. Are the responses voluntary, and if so, what are the problems with this in terms of inclusivity? I ask the students to think about these issues.

We talk about social statistics questions, such as voting. People may be excluded from the electorate because of their situation. How can you include people of no fixed abode when you need a local address to vote? What about someone who can’t prove their address because they’re staying with their auntie for six months?

I asked the students to consider a US survey of voting intentions carried out in 2024 by phoning people randomly. Who might be left out? We discussed how no one under 30 answers their phone if they do not know the caller, so this age group wouldn’t be included in the data. This example resonated with the students; it was clear how young people’s voices would be excluded from the data.

Appropriate questions

I ask students to consider which questions it is appropriate to ask in a survey. In displaced persons settlements (formerly called refugee camps), there’s a method of doing network mapping so that you can try to reunite people with their families, community, or neighbourhood when everyone gets re-homed. You ask people to name their friends because you want to reunite them. But it’s not appropriate to ask people to name their friends in some contexts, for example, if you are asking questions like these: ‘Who do you know who’s lying about working to receive benefits?’ ‘Who hasn’t paid their TV licence in your block of flats?’ We talk about why asking people to volunteer data can be unreliable from the data perspective, and we talk about designs that would be helpful and fair, and those that would be neither helpful nor fair.

Outcomes

Decolonisation of the curriculum

I reflect on thoughts and assumptions underlying knowledge and include diverse perspectives.

Student engagement

Since 10% of the assessment grade for this module is based on engagement, I allocate points after each lecture and each seminar. All students need to earn this mark, and there are different ways they can achieve it.

• ‘Fill in the blank’ activity: I start talking; they fill in their ideas. I create a space where students get credit for saying something – and it doesn’t matter if they’re right or wrong. Alternatively, they can ask a question.
• Students talk to the person next to them, using active listening and taking notes.
• They get credit for being present in the class and paying attention to us.
• In the computer seminars, we give students credit for helping the person next to them.

We also subtract points from students for negative behaviour, such as derailing conversations with personal agendas, criticising the efforts of other students, or staring at their phone for the entire session.

These methods encourage hesitant or shy people to speak. Students don’t shout out to gain points; there are no arguments; and no particular category of students dominates the discussion.

Comments

Student feedback has been excellent since I changed the module. Doctoral tutor Indrajit Adhikari describes his experience running the seminars:

Because students join the module with varied academic and professional backgrounds, we focus on creating a supportive environment where everyone can engage with quantitative methods at their own pace. We take care to use examples that avoid reinforcing stereotypes, and we explain the constraints of datasets – for instance, the difference between biological sex variables and gender identity.

In class, we occasionally use randomly generated datasets to demonstrate how statistical tools work. One example involved a hypothetical voting pattern that appeared to show female voters strongly favouring the far right and male voters leaning to the left. Because these patterns emerged purely from random data, they had no connection to real groups or real political behaviour.

We made a point of emphasising this to students because we wanted to avoid any implication that their biological sex was being linked to a particular political inclination. For many students, being associated with a far‑right or far‑left position solely on the basis of sex could feel uncomfortable, inaccurate, or even stigmatising.

By framing these results as illustrations rather than claims about real groups, we aim to create a respectful and inclusive learning environment. It reassures students that the examples are not reflections of our worldview or of actual demographic patterns, but simply tools for understanding statistical concepts.