Dr Michael Lewis

Dr Michael Lewis

School of Psychology

+44 (0)29 2087 5399
Tower Building, 70 Park Place, Cardiff, CF10 3AT
Media commentator
Available for postgraduate supervision

Research summary

Faces are highly important in human communication and interaction. My research explores various issues surrounding the psychology of seeing a face.

These issues include:

  • how we can distinguish between different faces so easily
  • how we perceive emotional expressions
  • how we detect faces are present in a scene
  • what happens when face recognition goes wrong such as in Capgras delusion.

I have also researched the effects of cosmetic therapies.

The techniques I employ include:

  • behaviour experimentation
  • patient studies
  • neural network and computational modelling, and brain imaging.

In addition, my research also explores methodological issues in other areas of cognitive psychology.

Undergraduate education

  • BSc First Class Hons in Mathematics and Psychology from The University of   Birmingham awarded in 1993.

Postgraduate education

  • My PhD was conducted in Cardiff University supervised by Robert Johnston and Hadyn Ellis. The thesis discussed and evaluated various models of face memory using a mixture of empirical studies and computational modelling. The degree was  awarded in 1998.

Professional memberships

  • British Psychological Society
  • Experimental Psychological Society.

Academic positions

  • Ocober 2011: promoted to Reader, Cardiff University
  • 1999-2011: Lecture/Senior Lecturer at Cardiff University
  • 1997-1999 Research Fellow at Cardiff University. Working with Hadyn Ellis on Capgras delusion.

Speaking engagements

My research has been featured several times in newspapers and on radio shows. Local BBC news also covered my research relating to cosmetic treatment.

Invited talks (eg University of Bristol, University of Essex, Lancaster University, Stirling University; Reading University)

Committees and reviewing

  • grant reviewing: BBSRC; ESRC; EPSRC; Leverhulme trust
  • journal reviewing (30 different journals, including JEP:Applied, JEP:HPP, JEP:LMC and Psychological Science)
  • BPS Cognitive Section Annual Conference 2010 organiser.


  • external examiner for MSc course in University of Kent
  • PhD examining (eg Lancaster University).






















I teach research methods at Level 1 covering issues related to experimental design.

I run a level 2 practical on facial feedback supporting PS2007.

I supervise a wide range of final year projects many relating to the psychology of the face.

Mixed-race faces

Quite often, black faces can all look the same to white people and the reverse effect can be observed for white people. How do mixed-race faces fit within this? Mixed-race faces offer interesting ways to look at how race is categorised by facial appearance. Further, genetic processes, such as hybrid vigour, have also been observed to affect mixed-race faces making them appear more attractive.

Lewis, M.B. (2010). Why are mixed-race people perceived as more attractive? Perception, 69, 136 – 138.

Botox and mood

Smiling makes us feel happy whereas frowning makes us feel sad. People who have undergone chemical denervation of their frown muscles (as in the treatment known as Botox) cannot frown. We have found that these people (possibly because they can no longer get the feedback from frowning) are happier than people for have had other forms of cosmetic treatment. Current research is exploring other possible psychological implications of these increasingly common cosmetic treatments.

Lewis, M.B. and Bowler, P.J. (2009). Botulinum toxin cosmetic therapy correlates with a more positive mood. Journal of Cosmetic Dermatology. 8, 24-26.

Navon effects

Reading the small letters of a large figure made out of small letters (a Navon figure) has unexpected effects. It makes people worse in face recognition tasks but it also makes wine recognition difficult.

My research tries to understand and apply these phenomena. For example, the same processes that cause this Navon effect might also explain why people are bad at recognising faces after they have been doing cryptic crosswords.

Lewis, M.B., Seeley, J. and Miles, C. (2009). Processing Navon letters can make wine taste different. Perception, 38, 1341-1346.

Lewis, M.B., Mills, C., Hills, P.J. and Weston, N. (2009). Navon letters affect face learning and face retrieval. Experimental Psychology, 56, 258-264.

Lewis, M.B. (2006). Last but not least: Eye-witnesses should not do cryptic crosswords prior to identity parades. Perception, 35, 1433-1436.

How can we spot a liar?

When people lie, do they first need to suppress the truth? If they do then can we use the additional time taken to do this to work out is someone is lying? Current research is looking at whether it takes longer to lie than to tell the truth. We are also exploring whether some people are particularly good at uncovering lies and trying to identify what cues they use.

Face detection

While much is known about factors that affect face recognition, relatively little is know about face detection – that is the ability to spot a face in a scene. This is surprising as it is a pre-requisite of recognition in the natural world. Modern technological devices (eg cameras) often come with face detection software but these are often fooled in ways that a human observer would not be. My research in this area represents the first systematic analysis of the psychological processes of face detection.

Lewis, M.B. & Edmonds, A. J. (2005). Searching for faces in scrambled scenes. Visual Cognition, 12, 1309-1336.

Lewis, M.B. & Edmonds, A.J. (2003). Face detection: Mapping human performance, Perception. 32, 903-920.

Statistical methods in psycholinguistics

The word ‘cat’ is read faster than ‘aardvark’ but is this difference because it is shorter, more frequent, learnt at an earlier age, has more similar words or just more furry? This is a question that has taxed psycholinguists for time without resolution. The problem with this field of research is that experimental research (that is where, for example, the length of the word ‘cat’ is manipulated) is impossible. My research aimed to address this issue using structural equation modelling – a method that tests hypothetical causative relationships between a variety of observational variables just as you might use to see whether socioeconomic status or race predicts performance in school.

Lewis, M.B. and Vladeanu, M. (2006). What do we know about psycholinguistic effects? Quarterly Journal of Experimental Psychology, 59, 977-986.

Lewis, M.B. (2006). Chasing psycholinguistic effects: A cautionary tale. Visual Cognition, 13, 1012-1026.

Models of face recognition

We recognise faces very quickly, but the difference in the time taken to recognise some faces allow us to generate models of how the information is stored. Typical (or average) faces are recognised more slowly than distinctive faces. A caricatured face can be recognised faster than an accurate image of a person. Faces of races that we are unfamiliar with are slower to be recognised. I developed and tested a model of face memory (called face-space-r) which was based on a few simple principles but could explain a wide range of patterns in how we recognise faces.

Lewis, M.B. (2004). Face-Space-R: Towards a unified account of face recognition. Visual Cognition, 11, 29-69.

Research collaborators

Internal collaborators:

  • Dominic Dwyer (Perceptual learning of faces)
  • Scott Jones (Perceptual learning of faces)
  • Lewis Bott (Detection of deceit)
  • John Patrick (Detection of deceit)
  • Emma Williams (Detection of deceit)
  • Chris Miles (Expertise and perception styles)
  • Rachael Elward (Face adaptation effects)
  • David Ross (Models of face recognition).

External collaborators:

  • Markus Bindemann (University of Kent; Face detection)
  • Peter Hills (Bournemouth University; Face recognition)
  • Patrick Bowler (Court House Clinics; Botox and facial feedback)
  • Nicola Weston (Police Sciences Unit; Face recognition)
  • Matei Vladeanu (Brunel University; Face and word recognition).

Faces are all very similar, yet we can distinguish among a large number of known faces and extract a wealth of information from them with remarkable speed and accuracy. My research explores this amazing ability from detecting a face in a visual scene, through recognition of a person, to retrieval of information concerning that person. I am also interested in the perception of emotional expressions, attractiveness and racial differences. While some experiments involve studying the learning of faces, other experiments explore facial illusions, caricatures, facial morphing or other visual manipulation of faces. Application of our understanding of face-related processes are important for studying dysfunctions of face recognition (eg, prosopagnosia, Capgras delusion) as well as within the forensic field.

If you are interested in applying for a PhD, or for further information regarding my postgraduate research, please contact me directly, or submit a formal application.

Current students

Scott Jones (Jointly supervised by Dominic Dwyer; funded by EPSRC). Scott is applying elements of perceptual learning in order to understand how faces move from being unfamiliar to familiar during learning. It is hoped that this understanding will lead to methods by which we can improve the learning of new faces.

Emma Williams (Jointly supervised by Lewis Bott; Funded by EADS). Emma is exploring the psychology of lying. She is investigating the cognitive effect of suppressing the truth when one lies. She is trying to identify whether some people are better at detecting lies and what cues they use to do this.

David Ross (Funded by ESRC). David is investigating the nature of our memory for faces. How are the many different faces that we can so quickly recognise stored?

Past projects

Peter Hills Peter tackled a wide range of face recognition related topics including adaption effects and other-race effects. He is currently a lecturer at Anglia Ruskin University.

Matei Vladeanu (Jointly supervised by Hadyn Ellis). Matei’s research aimed to understand the processes involved in priming in face recognition. This involved empirical research and neural network modelling. He is now a lecturer at Brunel University.

Orazio Guiffrida. Orazio analysed the lexical abilities of English/Italian bilinguials in order to better understand language processing. He is now a clinical psychologist in London.

Hazel Willis (took over supervision towards end of PhD). Hazel’s research lies in social cognition and intergroup relations. She is currently a lecturer at the University of Gloucestershire.

Laura Keyse (took over supervision towards end of PhD).

Sarah Sherwood (Jointly supervised by Hadyn Ellis).

Andrew Edmonds. Andrew studied the effect of inversion on face processing and recognition. Since completing his PhD, Andrew has held post-doctoral positions at the University of Kent and Nottingham Trent University.