Five lessons for using Mass Observation’s Covid-19 collections (plus five ways of solving the problem of representativeness)

One part of our project focuses on Covid-19 and the questions of responsibility outlined in our previous blog, but another part focuses on Mass Observation’s Covid-19 collections and questions of method. In the following blog, we outline some initial thoughts on using MO’s Covid-19 collections. At some point in 2021, we hope to organise a workshop where these and other thoughts can be discussed.

Five lessons for using Mass Observation’s Covid-19 collections (plus five ways of solving the problem of representativeness)

During 2020, Mass Observation (MO) collected lockdown diaries, 12 May day diaries, and responses to three letters/directives on life during the pandemic. These Covid-19 collections will be used by social scientists and historians to understand what ordinary people thought, felt, and did in Britain during 2020. But how should these collections be used? How should the material be read and interpreted? Drawing on histories of MO and existing research using MO sources, here are five lessons on how best to use MO’s Covid-19 collections.

1) Writing for MO should be treated as subjective writing. Writing for MOdoes a number of things. It transmits knowledge, constructs knowledge, and contests knowledge, but it tends to be subjective writing, as opposed to social reportage (Bloome et al 1993, Sheridan et al 2000). It tends to be writing in the service of identity construction (Nettleton and Uprichard 2011). It tends to be writing as a social and cultural practice oriented towards doing something: building a relationship, influencing an opinion, and especially positioning the author in relation to MO and other authorities and institutions (Sheridan et al 2000).

2) Writing for MO should be treated as dialogic. It is part of a dialogue with MO (Bloome et al 1993, Salter 2010). It is not just subjective, but intersubjective (Pollen 2014). The panellists are not just Mass Observers or respondents, but correspondents who exchange letters with MO as part of an ongoing conversation (Sheridan et al 2000). They have a sense of purpose and document ordinary life for an imagined audience of current and future researchers (Kramer 2014). Having said that, the specific wording of particular directives should not be overplayed. MO wording frames more than it dictates (Gazeley and Langhamer 2012).

3) Writing for MO provides evidence of top-down and bottom-up processes. In the writing of MO diarists, we see their cultural worlds – their worlds of discourse – and how they construct from these worlds their own distinctive selfhoods (Hinton 2008, 2010). So we see how individuals drive historical processes as historical agents operating in the face of received cultural norms (Hinton 2010). In writing for MO, we see evidence of top-down standards and codes, but also how people receive such standards and codes – their dis/comfort with them – and what people do with them (Langhamer 2016). In MO sources, we see the sociological constructs panellists use as resources, but also their lay articulations of those constructs – their practices of selection, interpretation, appropriation, incorporation, contextualisation (Wilson-Kovacs 2014).

4) MO not only reports public opinion, but also helps to shape it. The original MO of the 1930s and 1940s helped to develop new conceptions of population and culture, new notions of ‘mass’ (a particular conception of population) and ‘morale’ (a measure of the population’s mood), new means of knowing and governing the population, which it communicated to MO writers in the form of example diaries and directive responses, providing them with models for self-observation and so normalising mass surveillance as a collective habit (Harrison 2014). In its twentieth-century activities, which sometimes mobilised large numbers of people, MO benefited from and helped to produce a cadre of the technically, intellectually, scientifically engaged – a new social identity – and to embed social science in everyday life, providing people with a new language and imaginary of social group, social relationship, and social change (Savage 2008, 2010). MO provides participants with opportunities to tell their own stories, assert their own agency, reflexively fashion themselves (Sheridan 1993, Sheridan et al 2000). During the twentieth century, it contributed to the production of a new and more individualistic culture, which should not be interpreted as moral decline, but rather as the outcome of struggles for personal autonomy, egalitarianism, and democratisation (Hinton 2010).

5) The problem of representativeness can be addressed in at least five ways. The social constitution of MO’s panel of volunteer writers haunts social scientific engagements with MO data. The current version of the panel is more representative of the national population than previous versions, which were dominated by the radicalised lower middle class (Jeffrey 1978), or the technically-minded middle class (Savage 2010), or people from the middle classes, the Left, and London and South East England (Hinton 2013), or the elderly (Casey et al 2014). That being said, the current version remains unrepresentative in the statistical sense favoured by most social scientists, and it is dominated – as it always has been – by volunteers for a certain kind of project: self-selected enthusiasts who tend to be particularly dutiful, engaged, reflexive, and critical (Hinton 2010). How might users of MO’s Covid-19 collections respond to this problem of representativeness?

First, the skewed character of the panel can be challenged. Claims regarding the skewed character of the panel are often made on the basis of meta-data. For example, the panel is claimed to be skewed towards the middle classes because of the occupations registered by panellists at the time of joining. Such claims can be challenged. For example, Casey (2020) has argued that many panellists might be seen as middle class due to their current occupations, but might be seen as working class because that is how many socially mobile panellists continue to self-identify (evident in their writing, if not in the meta-data).

Second, the skewed character of the panel can be accepted and used to provide a focus for research. This may be a focus on women, who have often been over-represented on the panel, whose voices are often marginalised in public discourse, and who can be given voice by research using MO sources (Baker and Geringer 2018). Or it may be a focus on other groups often over-represented on the panel: the educated middle class (Savage 2007, 2008); the upwardly mobile (Casey 2020); particularly engaged, dutiful citizens (Manning 2017); volunteers, who tend to be elderly and female (Lindsey and Bulloch 2014); genealogists, who tend to be elderly, female, and middle class (Kramer 2011); gardeners, who tend to be elderly, female, middle class, and white (Bhatti et al 2009); and so on.

Third, the skewed character of the panel can be corrected, at least to some extent. Researchers can sample within the panel, filling quotas for age, gender, occupation, and place of residence (Clarke et al 2018, May 2018, Salter 2010). Clarke et al (2018) found that roughly 60 responses per directive allowed for wide social coverage and descriptive saturation.

Fourth, MO texts can be read carefully for what they tell us, without assuming they tell us about the views and experiences of representatives of particular social groups (the assumption behind the problem of representativeness). Here, a lead can be taken from the panellists themselves (Kramer 2014). Writing for an imagined audience of researchers, participants write reflexively and contextualise their responses, identifying the vantage points from which they write and explicitly positioning themselves as particular types of cases. Sometimes, they position themselves as typical cases (or typical of a particular social group). Sometimes, they position themselves as unusual (and, by doing so, describe what is more typical). This is the difference between ‘representative cases’ and ‘telling cases’ (Bloome et al 1993). The question to ask is not only ‘who speaks?’, but also ‘what do they tell us?’. Sometimes, MO sources make visible logical connections between phenomena of relevance far beyond the individual case or the particular social group that case might be assumed to represent (Clarke et al 2017). Something else MO sources tell us, when taken as a whole and not disaggregated as the views and experiences of representatives, is the range of cultural resources – categories, storylines, subject positions, folk theories – circulating in society at a particular historical moment (Clarke et al 2018; see also Gazeley and Langhamer 2012, Nettleton and Uprichard 2011, Salter 2010, Savage 2007). It is from these cultural resources that people construct understandings. To establish the full range, a horizontal approach to the archive – quoting from as many different writers as possible – is appropriate (Clarke et al 2018). However, to establish how these cultural resources are used, contested, and transformed (see Lesson 3), a more vertical approach is recommended, where particular panellists are followed across multiple directives (e.g. Busby 2000).

Fifth, in the spirit of the original MO, the problem of representativeness can be replaced with a similar but different problem: the problem of representation. The founders of MO in the late 1930s were interested in how to represent everyday life and approached this question not as a statistical problem focused on inputs – a sampling problem – but as a literary problem of composition and depiction (Highmore 2002, Hubble 2010, Jardine 2018). Like the avant-garde continental sociologists laying the groundwork for everyday life studies at the time (Simmel, Benjamin etc.), or the scientific humanists of the time (Tarde, Freud etc.), Charles Madge and Humphrey Jennings especially were influenced by surrealism and the aesthetic techniques of surrealist poetry, painting, film, and theatre. They sought images, metaphors, close-ups, luminous moments, multiple points of view, which they thought provided clues to the mass unconscious, and which they arranged in montages to draw unfamiliar associations, to emphasise ambiguity and possibility, to represent society as a diverse, heterogeneous ‘totality of fragments’ (Highmore 2002), and to promote a new consciousness and encourage social transformation (at a time of concerns about mass media, elite representations of a unified ‘people’, and the rise of Nazism in Germany).

So where does all this lead in terms of research questions that might be asked of MO’s Covid-19 collections? We could ask: what cultural resources, worlds of discourse, constructs, categories, storylines, subject positions, norms, standards, codes, and theories were people exposed to during the pandemic? How did people receive, use, and transform these resources – by means of selection, interpretation, articulation, contestation, appropriation, incorporation, contextualisation? What was produced by these processes – what new concepts, notions, languages, models, habits, imaginaries, identities, selves, cultures, ways of knowing, ways of governing? How best to represent all this aesthetically, in a way that depicts the totality of British society (the ‘mass’ of MO), but also captures fully the diversity, heterogeneity, ambiguity, and possibility of life under/after the pandemic?


Baker J and Geringer D (2018) ‘Space, text, and selfhood: Encounters with the personal computer in the Mass Observation Project archive, 1991-2004’, Contemporary British History 33(3): 293-312.

Bhatti M, Church A, Claremont A, and Stenner P (2009) ‘‘I love being in the garden’: Enchanting encounters in everyday life’, Social and Cultural Geography 10(1): 61-76.

Bloome D, Sheridan D, and Street B (1993) ‘Reading Mass-Observation writing: Theoretical and methodological issues in researching the Mass-Observation archive’, Mass-Observation Archive Occasional Paper 1 (University of Sussex Library).

Busby H (2000) ‘Writing about health and sickness: An analysis of contemporary autobiographical writing from the British Mass-Observation Archive’, Sociological Research Online 5(2).

Casey E (2020) ‘Gambling, status anxiety, and inter-generational social mobility: Findings from the Mass Observation Archive’, Sociology 54(2): 380-396.

Casey E, Courage F, and Hubble N (2014) ‘Special section introduction: Mass Observation as method’, Sociological Research Online 19(3).

Clarke N, Jennings J, Moss J, and Stoker G (2017) ‘Changing spaces of political encounter and the rise of anti-politics: Evidence from Mass Observation’s general election diaries’, Political Geography 56: 13-23.

Clarke N, Jennings W, Moss J, and Stoker G (2018) The Good Politician: Folk Theories, Political Interaction, and the Rise of Anti-Politics (Cambridge, Cambridge University Press).

Gazeley I and Langhamer C (2012) ‘The meanings of happiness in Mass Observation’s Bolton’, History Workshop Journal 75(1): 159-189.

Harrison R (2014) ‘Observing, collecting, and governing ‘ourselves’ and ‘others’: Mass-Observation’s fieldwork agencements’, History and Anthropology 25(2): 227-245.

Highmore B (2002) Everyday Life and Cultural Theory: An Introduction (London, Routledge).

Hinton J (2008) ‘‘The ‘class’ complex’: Mass-Observation and cultural distinction in pre-war Britain’, Past and Present 199: 207-236.

Hinton J (2010) Nine Wartime Lives (Oxford, Oxford University Press).

Hinton J (2013) The Mass Observers: A History, 1937-1949 (Oxford, Oxford University Press).

Hubble N (2010) Mass Observation and Everyday Life: Culture, History, Theory (Basingstoke, Palgrave Macmillan).

Jardine B (2018) ‘Mass-Observation, surrealist sociology, and the bathos of paperwork’, History of the Human Sciences 31(5): 52-79.

Jeffrey T (1978) Mass Observation – A Short History (Birmingham, University of Birmingham).

Kramer A-M (2011) ‘Kinship, affinity, and connectedness: Exploring the role of genealogy in personal lives’, Sociology 45(3): 379-395.

Kramer A-M (2014) ‘The observers and the observed: The ‘dual vision’ of the Mass Observation Project’, Sociological Research Online 19(3).

Langhamer C (2016) ‘An archive of feeling? Mass Observation and the mid-century moment’, Insights 9(4): 1-15.

Lindsey R and Bulloch S (2014) ‘A sociologist’s field notes to the Mass Observation Archive: A consideration of the challenges of ‘re-using’ Mass Observation data in a longitudinal mixed-methods study’, Sociological Research Online 19(3).

Manning N (2017) ‘The veil of duty: Can dutiful forms of citizenship mask feelings of political dissatisfaction?’, British Politics 13: 467-483.

May V (2018) ‘Belonging across the lifetime: Time and self in Mass Observation accounts’, The British Journal of Sociology 69(2): 306-322.

Nettleton S and Uprichard E (2011) ‘‘A slice of life’: Food narratives and menus from mass-observers in 1982 and 1945’, Sociological Research Online 16(2).

Pollen A (2014) ‘Shared ownership and mutual imaginaries: Researching research in Mass Observation’, Sociological Research Online 19(3).

Salter A (2010) ‘Filling the silences? Mass-Observation’s wartime diaries, interpretive work, and indexicality’, Life Writing 7(1): 53-65.

Savage M (2007) ‘Changing social class identities in post-war Britain: Perspectives from Mass-Observation’, Sociological Research Online 12(3).

Savage M (2008) ‘Affluence and social change in the making of technocratic middle-class identities: Britain, 1939-55’, Contemporary British History 22(4): 457-476.

Savage M (2010) Identities and Social Change in Britain since 1940: The Politics of Method (Oxford, Oxford University Press).

Sheridan D (1993) ‘Writing to the archive: Mass-Observation as autobiography’, Sociology 27(1): 27-40.

Sheridan D, Street B, and Bloome D (2000) Writing Ourselves: Mass-Observation and Literary Practices (New Jersey, Hampton Press).

Wilson-Kovacs D (2014) ‘‘Clearly necessary’, ‘wonderful’, and ‘engrossing’? Mass Observation correspondents discuss forensic technologies’, Sociological Research Online 19(3).