How do you measure a feeling?

Tom Jenkins won the 2021 John Grace QC PhD Scholarship and is carrying out his research at Bath University. Tom’s project is titled ‘Developing our Understanding of Dehumanisation in People with Psychosis’ and below he writes about what he learned from a course that he recently attended in support of his PhD work.

When planning for my PhD, I was curious about the impact of dehumanisation – feeling less or other than human – in people with psychosis. I wanted to measure dehumanisation but found there is no well-developed way in which to do this. In fact, there are even conflicting ideas about how dehumanisation is experienced. I was confused by this, and began to wonder - what is a feeling? And how do we measure one?

An effective way to measure feelings is by questionnaire. Questionnaires contain a list of ‘items’ (statements) and are completed by indicating how much one agrees or disagrees with each. The responses provide a score which represents feelings of a ‘construct’ (area of interest). In our case, this construct is dehumanisation. Rigorous questionnaire development takes time and involves several stages. I am very thankful to MHRUK for funding my attendance to a training course which taught this methodology, and I am writing this blog to share what I have learned.

The process begins with a literature review – a thorough reading of all theories, definitions, and previous measures of the construct. Alongside, people from the target population are interviewed about their experiences of the construct. In our project, we have conducted a literature review on dehumanisation, and spoken to people with psychosis about their experiences of dehumanisation. Items are generated from the main themes of the literature review and interview responses. If, for example, ‘sense or loss of belonging’ is a theme that emerges, an item like ‘I feel a sense of belonging with others’ may be generated. Included in the development is a team of experts, which for us is people with lived experience of psychosis, clinicians, and academics. Once we have a draft of the questionnaire, another set of interviews is conducted to get feedback on whether the items make sense. The merit of great questionnaire development is the inclusion of a diverse group of people and the integration of their perceptions and expertise.

The questionnaire will then be given to a large sample of people and statistical tests are conducted to eliminate items that are less relevant, making the questionnaire more concise and user-friendly. It is also assessed for reliability and validity. Reliability is the ability of the questionnaire to produce the same results consistently. One way to test this would be to administer the questionnaire to a sample of people twice in quick succession and see if the scores are similar. Validity is the ability of the questionnaire to measure the construct accurately. One way to test this is to see whether scores on the questionnaire are similar to scores on related measures. In our case, we might expect that a person who feels dehumanised may also feel stigmatised. Once the checks are complete, the questionnaire will be finished with the assurance that what is intended to be measured is what is actually being measured.

In our study, we hope that measuring feelings of dehumanisation will help us understand more about the impact that psychosis can have on mental health. This will inform thinking for how psychological therapy can be adapted to help people to reconnect with their sense of humanity.

Tom Jenkins