Empirical Sociology: Studying Society Through Evidence

Empirical sociology is the study of social life through the collection, processing and communication of primary data: observations, counts, interviews, surveys, documents and records gathered from the world rather than deduced from first principles. Its subject matter is as wide as society itself. Empirical sociologists study the economy, law, family life, religion, politics, cities, work, crime and health, and they often treat everyday common sense as a form of knowledge worth examining in its own right.

The phrase can mean two slightly different things. In a broad sense, it describes any sociology that rests on evidence, as opposed to purely theoretical or philosophical reflection about society. In a narrower, historical sense, it refers to particular traditions of research, especially the American tradition that grew out of the social reform movements of the Progressive Era and the University of Chicago, and the European traditions of social surveys and "sociography". This article covers both: what empirical sociology is, where it came from, how it works in practice, the landmark studies that shaped it, the debates it has provoked, and its current challenges.

Definitions and key ideas

Different writers have put the idea in different ways. The German sociologist Ferdinand Tönnies described empirical sociology as the collection of factual information or evidence about society. The American sociologist Morris Janowitz described it as the collection and planning of empirical indicators of social change, and he drew a close connection between sociology and history, writing that "All empirical sociology is history, but history is not all empirical sociology." His point was that every piece of social data is a record of a particular time and place, but that history includes much more than the systematic measurement of change.

Three related ideas are often confused with empirical sociology, and it helps to separate them.

Empirical sociology is often lumped together with positivism, especially by its critics. Theodor Adorno and other members of the Frankfurt School attacked what they saw as positivist sociology, and they also criticised Talcott Parsons's systems theory. But empirical sociology is broader than positivism. A researcher who spends a year living in a neighbourhood and writing detailed field notes is doing empirical work, even if they have no interest in discovering universal laws.

Historical roots

Social surveys in nineteenth-century Britain

Long before sociology existed as a university subject, reformers, officials and philanthropists were counting and mapping social conditions. One of the most influential was the London shipowner Charles Booth. His survey of life and labour in London ran from 1886 to 1903. Booth's investigators classified streets and households into eight classes, labelled A to H. Classes A to D represented various degrees of want, and classes E to H represented comfort.

Booth's maps became famous. A pilot study of the East End included "A Descriptive Map of London Poverty in 1889", hand-coloured and printed by the Ordnance Survey, using colours to show levels of poverty street by street. A map covering the whole city, the "Map shewing Degrees of Poverty in London", was published in 1891 and showed data collected in 1889 and 1890. The full work appeared in 1902 and 1903 as the seventeen volumes of Life and Labour of the People in London; earlier parts had been published in 1889 and 1891. A New Zealand newspaper report from 1894 summarised Booth's earlier finding that 30.7 per cent of the London population he studied were living in poverty and 69.3 per cent in comfort. That figure is widely repeated, though it should be read as an estimate based on his classification scheme rather than a modern poverty measure.

Booth's method, combining house-to-house information with maps and classification, became a model for social investigators in Britain and abroad.

Settlement houses and the Progressive Era in the United States

Empirical sociology in the American sense has strong roots in the social reform movements of the Progressive Era, which ran roughly from the 1890s to the 1920s. Settlement houses, where middle-class reformers lived among poor and immigrant communities, were important sites of early research.

In 1895, residents of the Hull-House settlement in Chicago, including Florence Kelley and Jane Addams, published Hull-House Maps and Papers. It presented the nationalities and wages of people living in a congested district of Chicago, together with essays on social conditions. Like Booth's work, it used maps to make social facts visible. The study drew attention to poverty and discrimination in the neighbourhood and helped build support for social services and legislation. The University of Illinois Press reissued it in 2007.

In 1896 the scholar W. E. B. Du Bois arrived in Philadelphia, invited in connection with the settlement movement, to study the city's Black population. The result, The Philadelphia Negro (1899), focused on the Seventh Ward, a neighbourhood with a large African American population. Du Bois carried out a door-to-door inquiry over about fifteen months, gathering information on family status, occupations, religion and other aspects of life, and combined it with historical, economic and social data. Social survey research was still in its infancy, and Du Bois did not begin with a ready-made set of methods; he largely developed them as he went. One source says he used Booth's London survey and Hull-House Maps and Papers as models. The book has been described as the first scholarly study of race in an American city, and later scholarship has praised it particularly for its empirical and methodological quality. For a long time, though, it was grouped with "social realist" writing about poverty and its sociological significance was underestimated.

The Chicago School

The University of Chicago founded the first independent department of sociology in the United States in 1892, under Albion Small, who also established the American Journal of Sociology in 1895 and co-wrote what has been described as the first sociology textbook with George E. Vincent in 1894. The department became the centre of advanced sociological thought in the United States between about 1915 and 1935, and the group of researchers there is known as the Chicago School. The Wikipedia article on empirical sociology describes the Chicago School's work between 1895 and 1929 as marking the first stage of empirical sociology, because it introduced changes to scientific procedure.

What made Chicago distinctive was its insistence on studying the city around it at first hand. Key figures included Robert E. Park, Ernest Burgess, Roderick McKenzie, W. I. Thomas, Louis Wirth and Edwin Sutherland. Their methods included:

• Mapping official data. Researchers plotted census, housing, welfare and crime records on maps. Clifford Shaw and Henry McKay, for example, produced spot maps, rate maps and zone maps of delinquency across Chicago.
• Life histories and personal documents. W. I. Thomas pioneered the use of letters, diaries and autobiographies as sociological evidence.
• Fieldwork. Researchers went into neighbourhoods, dance halls, gangs and immigrant communities to observe and interview.

From this work came theories that are still taught. Burgess's concentric zone model, set out in The City (1925) by Park, Burgess and McKenzie, described the city as growing outward in rings from a central business district, through a "zone in transition", to workers' homes, residential areas and finally commuter zones. The idea of social disorganisation, which links neighbourhood problems to a breakdown in local institutions and relationships, was another product of Chicago's empirical work.

The Polish Peasant (1918 to 1920)

The most ambitious early Chicago study was The Polish Peasant in Europe and America by W. I. Thomas and Florian Znaniecki, published in five volumes between 1918 and 1920. It studied Polish immigrants to the United States and their families in Poland, drawing mainly on personal documents. Volume I contains roughly fifty series of letters, more than 700 in all, sorted into family, marital and other categories. Volume II includes the autobiography of a single immigrant, Władysław Wiszniewski, known as Wladek, which one reviewer described as a 312-page life history. It is often credited as the first systematically collected sociological life history. Scholars still debate how the work was divided between the two authors; one account says the idea originated with Thomas while Znaniecki drafted much of the text.

The study mattered not just for its findings about migration and family change but for its demonstration that ordinary people's own words could be treated as serious data.

Middletown (1929)

In 1924 Robert and Helen Lynd moved to Muncie, Indiana, to study daily life in a single Midwestern community over about eighteen months. They published Middletown: A Study in Modern American Culture in 1929, and after returning to Muncie in 1935 they published Middletown in Transition (1937), which analysed how the Great Depression had changed the town. The first book is often described as a pioneering community study. Later critics pointed out a significant limitation: the Lynds' study included only Muncie's white residents. That gap is a reminder that empirical research is shaped by choices about whom to study, and those choices can leave whole groups out of the picture.

Empirical sociology in Europe

The Wikipedia article stresses that empirical sociology also has important European roots. The British sociologist Barry Smart, in a 2013 book, argued that empirical sociology became established in Eastern Europe. The Polish-Austrian sociologist Ludwig Gumplowicz researched society and law with a focus on conflict between groups of people. The Austrian philosopher of science Otto Neurath, a leading member of the Vienna Circle of logical positivists and the main author of its manifesto, wrote a 1931 work on "empirical sociology" and argued that the social sciences should be grounded in statements about observable, physical events. Neurath also founded a museum of society and economy in Vienna (established in 1923 and renamed the Gesellschafts- und Wirtschaftsmuseum in 1925) and developed the Isotype system of pictorial statistics to communicate social data to the public.

One of the best-known European empirical studies is Die Arbeitslosen von Marienthal (The Unemployed of Marienthal). In the winter of 1931 to 1932, a team including Marie Jahoda, Paul Lazarsfeld and Hans Zeisel studied Marienthal, a small industrial village near Vienna, after its main factory closed. About half the workers were jobless and around three-quarters of families were affected. The team studied the community as a whole rather than isolated individuals, combining quantitative procedures with qualitative observation of families. The book, published in 1933, found not revolutionary anger but apathy: a reduced sense of expectation and activity, a disrupted sense of time, and what the study called a "tired community". The finding challenged both the left-wing expectation that unemployment would radicalise workers and the market view that cutting welfare would spur people into action. Jahoda wrote the main text, Zeisel the sociographic appendix and Lazarsfeld the introduction. Lazarsfeld emigrated to the United States in 1935, where he became a leading figure in survey and communications research.

The rise of survey research

From the 1930s onward, the sample survey became the signature tool of empirical sociology, especially in the United States. Instead of studying every household, as Booth had tried to do, researchers could study a carefully chosen random sample and use probability theory to estimate how closely the results matched the whole population.

Large, repeated surveys followed. The General Social Survey (GSS), run by NORC at the University of Chicago, began in 1972 as an in-person survey funded by the National Science Foundation, which remains its main funder, and the Russell Sage Foundation. It ran annually until 1994, except in 1979, 1981 and 1992, and has been conducted in even-numbered years since then. Because it repeats many of the same questions in the same wording, it has become one of the most important sources for tracking changes in American attitudes over time. In Europe, the European Social Survey (ESS) began its first round in 2002 to 2003 and has been held every two years since, with a core questionnaire that stays relatively constant and rotating modules on particular topics.

How empirical sociology works

Empirical sociology is less a single method than a family of methods. They can be grouped by the kind of data they produce.

The research cycle

Most empirical studies, whatever their method, move through a similar cycle:

1. Formulate a question. For example: does long-term unemployment make people more politically active or less?
2. Define concepts and measures. What counts as "unemployed"? How will "political activity" be measured: voting, attending meetings, reading newspapers?
3. Choose a design. A survey of many people, an in-depth study of one community, a comparison between places, a study over time?
4. Select cases or a sample. Whom or what will be studied, and how will they be chosen?
5. Collect data. Through questionnaires, observation, interviews, records or other sources.
6. Analyse. Count, compare, code and interpret.
7. Communicate. Write up results, including limitations, so others can evaluate and build on them.

The Wikipedia article stresses that communication is part of the definition. Data that are collected but never shared, or shared without explanation of how they were gathered, do little to advance knowledge.

Quantitative and qualitative approaches

A long-running division in empirical sociology is between quantitative work, which deals in numbers and statistical relationships, and qualitative work, which deals in words, meanings and detailed description. In practice, many of the most famous studies combine both. Marienthal used counts and measurements alongside close observation of families. Du Bois used statistics and maps alongside historical and descriptive analysis. Modern "mixed methods" research continues this tradition.

Landmark studies and what they showed

Durkheim's Suicide (1897)

The French sociologist Émile Durkheim's Le Suicide, published in 1897 and translated into English in 1952, is often treated as a founding work of empirical sociology. Durkheim used official statistics from several European countries to argue that suicide rates, which seem to be the result of deeply personal decisions, vary systematically with social conditions. He proposed four types:

Durkheim argued, for example, that Catholics had lower suicide rates than Protestants because Catholic communities exercised stronger social control. The study has been heavily criticised. A common objection is that it commits the ecological fallacy: drawing conclusions about individuals from data about groups. Showing that a region with more Protestants has a higher suicide rate does not prove that the Protestants themselves are the ones taking their lives. Other scholars have defended Durkheim against this charge, and the debate continues. Either way, Suicide showed that social patterns could be detected in aggregate data, and it set the template for using official statistics in sociology.

The Hawthorne studies (1924 to 1932)

Between 1924 and 1932, researchers studied productivity at Western Electric's Hawthorne Works near Chicago. The work included experiments on lighting, the relay assembly test room, a large interview programme and the bank wiring observation room. The studies became famous for suggesting that social factors, including the attention of researchers, affected productivity as much as physical conditions. The idea that being observed changes behaviour became known as the "Hawthorne effect".

Later reanalysis has complicated this story. In a 2009 working paper for the National Bureau of Economic Research, the economists Steven Levitt and John List recovered data from the original illumination experiments and concluded that they did not provide strong evidence for the effect as usually described, while leaving room for subtler effects. In the bank wiring room, intensive observation did not produce a productivity increase; instead, workers held output down through informal group norms. The Hawthorne studies remain important, both for what they revealed about informal work groups and as a cautionary tale about how a memorable finding can outrun its evidence.

Robert K. Merton and middle-range theory

The American sociologist Robert K. Merton developed theories of deviance and social roles that drew on and guided empirical research. Merton is associated with the idea that sociology should build theories of the "middle range", more general than single descriptions but more testable than grand systems of thought. This was, in part, an attempt to bridge the gap between empirical research and abstract theory.

Debates and criticisms

Empirical sociology has never been without critics. The arguments are worth understanding because they shaped how the field developed.

Grand theory versus abstracted empiricism

In The Sociological Imagination (1959), the American sociologist C. Wright Mills attacked two tendencies he saw in American sociology. One was "grand theory", represented by Talcott Parsons, which in Mills's view treated society in pure abstraction without attention to history or real social context. The other was "abstracted empiricism", represented by Paul Lazarsfeld, which Mills saw as treating sociology mainly as a set of methodological tools. Mills argued that this kind of research tended toward psychologism, explaining behaviour at the level of individuals without attention to social structure, and that it separated research from theory. Later commentators have summarised the complaint as putting method above substance.

Mills's critique was not a rejection of evidence. His argument was that good sociology needed to connect individual experience to history and social structure, and that piling up data without asking big questions was as unhelpful as theorising without data.

The positivism dispute

In Germany, a famous debate known as the Positivismusstreit (positivism dispute) began at the 1961 meeting of the German Sociological Association in Tübingen. On one side were the critical rationalists Karl Popper and Hans Albert; on the other, Theodor Adorno and Jürgen Habermas of the Frankfurt School. The Frankfurt School accused their opponents of positivism, while Popper and Albert regarded themselves as opponents of positivism. The debate built on an earlier controversy over whether social science could or should be value-free, and the essays were collected and published in 1969. The dispute is a reminder that "empirical" and "positivist" are not the same thing, and that the question of what evidence can and cannot tell us about society is philosophical as well as technical.

Ethics and the treatment of research subjects

Empirical research involves real people, and some studies have crossed ethical lines. Laud Humphreys's Tearoom Trade (1970), a study of anonymous sexual encounters between men in public restrooms, is one of the most discussed examples. Humphreys acted as a lookout without revealing that he was a researcher, and later carried out follow-up interviews with some of the men in their homes without seeking their informed consent; those interviews drew the harshest criticism. Critics argued he had deceived his subjects and violated their privacy. Humphreys himself later said he should have identified himself as a researcher. Some commentators have also noted that part of the outrage was a hostile reaction to the subject matter itself.

In the United States, a national commission produced the Belmont Report, published in the Federal Register on 18 April 1979, which set out three principles for research involving human subjects: respect for persons, beneficence and justice. Respect for persons underlies the requirement for informed consent; beneficence requires that research be justified by a favourable balance of risks and benefits; justice concerns the fair selection of research subjects. These principles shape the work of institutional review boards that now approve sociological studies at universities.

Representativeness and response rates

Survey research depends on reaching a sample that represents the population. That has become harder. Pew Research Center reported in 2019 that response rates to typical telephone surveys had held at about 9 per cent for several years before falling to 7 per cent in 2017 and 6 per cent in 2018. Pew also noted that its own studies in 1997, 2003, 2012 and 2016 found little relationship between response rates and accuracy, and that weighting can correct some biases. But it warned that low response increases the risk of error when the attitudes being measured also affect who agrees to take part. Its example was volunteering, a topic where standard weighting falls short and produces biased estimates.

Big data and computational social science

In 2009, a group of fifteen researchers led by David Lazer published a short article in Science titled "Computational Social Science". It argued that a new field was emerging that could collect and analyse data at a scale capable of revealing patterns of individual and group behaviour. Digital traces from phones, social media, online purchases and other sources now allow sociologists to observe behaviour directly and at enormous scale.

These data raise familiar empirical problems in new forms. The people who use a particular platform are not a random sample of society. The data are usually collected by companies for their own purposes and may not measure what researchers want. Access can be withdrawn at any time. Privacy and consent are far harder to manage when data are collected passively. The core lessons of empirical sociology, such as asking who is in the data, who is left out, and how the measures were constructed, are as relevant to big data as they were to Booth's street surveys.

Worked examples (hypothetical)

The following examples are invented to illustrate common calculations and reasoning. They are not results from real studies.

Example 1: The margin of error of a survey

Suppose a researcher surveys a simple random sample of 1,000 adults and finds that 50 per cent agree with a statement. How precise is that estimate?

For a proportion, the standard error is the square root of p multiplied by (1 minus p), divided by the sample size. With p = 0.5 and n = 1,000:

• p × (1 − p) = 0.5 × 0.5 = 0.25
• 0.25 ÷ 1,000 = 0.00025
• The square root of 0.00025 is about 0.0158

At the conventional 95 per cent confidence level, the margin of error is about 1.96 times the standard error: 1.96 × 0.0158 ≈ 0.031, or about 3.1 percentage points. The researcher could report the result as 50 per cent, plus or minus about 3 points.

Two cautions apply. First, this calculation assumes a true random sample. If the people who respond differ systematically from those who do not, the real error can be much larger than the margin of error suggests. Second, the margin of error for subgroups is larger, because the sample size is smaller. If only 200 of the 1,000 respondents are under 30, the margin of error for that group at p = 0.5 is about 1.96 × the square root of (0.25 ÷ 200), which is about 1.96 × 0.0354, or roughly 6.9 points.

Example 2: Calculating a response rate

A research team selects 10,000 telephone numbers at random and completes interviews with 600 people. In the simplest calculation, the response rate is 600 ÷ 10,000 = 6 per cent. (Professional survey organisations use more detailed formulas that account for numbers that are not working or not eligible, but the basic idea is the same.) A 6 per cent response rate does not automatically mean the results are wrong, but it does mean the researchers need to check carefully whether the 600 who answered differ from the 9,400 who did not.

Example 3: The ecological fallacy

Imagine two districts. District A has a high proportion of older residents and a high rate of a particular illness. District B has a younger population and a low rate. It might be tempting to conclude that older people are more likely to have the illness. But the district-level data alone cannot show this. It could be that the illness in District A is concentrated among younger residents working in a local industry, while the older residents are healthy. To make claims about individuals, researchers need data about individuals. This is the logic behind the criticism of Durkheim's Suicide.

A timeline of empirical sociology

Designing a small empirical study: a practical guide

Readers who want to try empirical work themselves, for a class project or a community question, can follow the same logic used by professional researchers. The steps below describe good practice in general terms; they are not tied to any particular study.

Start with a question that evidence can answer. "Is my town a good place to live?" is too vague. "How do residents of my neighbourhood rate local public transport, and does this differ between people who own cars and people who do not?" can be studied.

Decide what you will measure, and write it down. If you are asking about transport, decide whether you will ask people to rate it on a scale, ask how often they use it, record how long journeys take, or all three. Writing definitions down before collecting data protects against the temptation to change them once you see the results.

Think hard about who you will reach. Standing outside a railway station will reach people who use trains, not those who have given up on them. A convenience sample can still be useful, but its limits should be stated honestly. This is the same issue that troubles national surveys with low response rates.

Combine methods where you can. A short questionnaire might show that people without cars rate transport poorly. A handful of conversations might reveal why: perhaps the last bus leaves too early for shift workers. Numbers show patterns; words help explain them.

Respect the people you study. Tell them who you are and what you are doing, let them decline, and do not record names or details you do not need. These are small-scale versions of the principles in the Belmont Report.

Report what you found and how you found it. Include the questions you asked, how many people took part, how they were chosen and what you could not find out. Others can only trust and build on your findings if they can see how they were produced, which is why communication is part of the very definition of empirical sociology.

Applications

Empirical sociology is not only an academic exercise. Its methods and findings feed into many areas of public life.

• Social policy. From Booth's poverty maps to modern surveys of income and living conditions, empirical research has shaped how governments understand and respond to poverty, housing, health and education.
• Urban planning. The Chicago School's maps of cities influenced how planners thought about neighbourhoods, and spatial analysis of social data remains central to planning today.
• Labour and employment. Studies such as Marienthal and the Hawthorne research shaped understanding of unemployment and workplace behaviour.
• Public opinion. Repeated surveys such as the GSS and ESS track changes in attitudes on topics from religion to trust in institutions, providing evidence for public debate.
• Criminology. Empirical studies of crime and delinquency, from Shaw and McKay's maps onward, have influenced policing and crime prevention.
• Civil rights and inequality. Du Bois's Philadelphia study is an early example of evidence used to document racial inequality, and empirical research remains central to debates about discrimination.
• Business and markets. Survey methods and interview techniques developed in sociology are used widely in market research and organisational studies.

Limits of empirical sociology

Empirical sociology has achieved a great deal, but it has real limits that are worth stating plainly.

• Measurement shapes findings. How a concept is defined and measured affects what is found. Two surveys that define "poverty" differently can reach different conclusions from the same population.
• Correlation is not causation. Many empirical studies show that two things go together but cannot show that one causes the other. Experiments and careful research designs help, but many social questions cannot be studied experimentally.
• Observation can change behaviour. As the Hawthorne research, even in its revised form, reminds us, people may act differently when they know they are being studied.
• Data reflect institutions. Official statistics record what institutions choose to record. Crime figures, for example, reflect policing practices as well as crime itself.
• Who is left out. Studies can exclude groups through their design, as the original Middletown study did. Low response rates and unrepresentative digital data create new forms of exclusion.
• Values are not removed by data. The positivism dispute and Mills's critique both show that choices about what to study and how to interpret findings involve values and theory, not just evidence.

Common misconceptions

"Empirical sociology means quantitative sociology." It does not. Ethnography, interviews and document analysis are empirical methods. Many classic studies combined numbers and words.

"Empirical sociology has no theory." Empirical research always involves theoretical choices about what to look for and how to interpret it. Durkheim's Suicide, for example, is both an empirical study and a theoretical argument about social integration and regulation.

"Empirical sociology is the same as positivism." Positivism is one philosophical view of what social science should be. Many empirical sociologists reject strict positivism while still insisting on evidence.

"A big sample guarantees accuracy." A large sample reduces random error but does nothing about systematic bias. A very large but unrepresentative dataset can be more misleading than a smaller, well-designed one.

"The Hawthorne effect is a proven law." The original studies are now viewed more cautiously. The idea that observation can change behaviour is reasonable and sometimes supported, but the Hawthorne studies themselves are weaker evidence for it than popular accounts suggest.

"Empirical sociology began at Chicago." Chicago was crucial for American sociology, but social surveys, mapping and statistical studies in Britain, France, Germany, Austria and Eastern Europe were also central to the field's development.

Frequently asked questions

What is the difference between sociology and empirical sociology?

Sociology is the study of society as a whole, including theory, history and philosophy of social life. Empirical sociology is the part of sociology that gathers and analyses evidence. In practice, most sociologists do some empirical work, and most empirical sociologists use theory.

Who were the first empirical sociologists?

There is no single answer. Charles Booth in London, the residents of Hull-House in Chicago, W. E. B. Du Bois in Philadelphia, Émile Durkheim in France and the Chicago School researchers are all frequently named as early pioneers. Each represents a different strand: social surveys, settlement research, statistical analysis and urban fieldwork.

Why did the Chicago School matter so much?

The University of Chicago had the first independent sociology department in the United States and the first American sociology journal. Its researchers insisted on studying the city at first hand, developed new methods such as the systematic use of life histories and social mapping, and produced theories, such as the concentric zone model and social disorganisation, that shaped the field for decades.

Is survey research still reliable?

Surveys remain a central tool, but their reliability depends on design. Response rates have fallen, which makes careful sampling and weighting more important. Pew Research Center has found that low response rates do not by themselves make surveys inaccurate, but they increase the risk of bias when the topic being measured is linked to willingness to respond.

What is the ecological fallacy?

It is the error of drawing conclusions about individuals from data about groups. If a city with many retired people has a high rate of a particular behaviour, that does not by itself show that retired people are the ones engaging in it.

How do sociologists protect the people they study?

Most universities require research involving human subjects to be approved by an ethics committee or institutional review board. In the United States, the principles of the 1979 Belmont Report (respect for persons, beneficence and justice) guide these reviews. Informed consent, confidentiality and minimising harm are standard requirements.

Is big data replacing traditional empirical sociology?

Not replacing, but adding to it. Large digital datasets allow new kinds of research, but they bring their own problems of representativeness, access and privacy. Surveys, interviews and fieldwork remain essential, especially for understanding what behaviour means to the people involved.

Why empirical sociology matters

Empirical sociology rests on a simple but demanding idea: that claims about society should be checked against evidence gathered carefully from society itself. From Booth's coloured maps of London streets to the nationally representative surveys of today, empirical sociologists have tried to replace guesswork, prejudice and anecdote with systematic observation.

The history of the field also shows that evidence does not interpret itself. Durkheim's statistics needed theory to make sense; the Hawthorne findings needed reanalysis decades later; Middletown's picture of American life left out people who did not fit the researchers' frame. Good empirical sociology combines careful data collection with honest reflection about what the data can and cannot show. That combination is as valuable now, in an age of digital data and falling survey response rates, as it was when reformers first walked from door to door counting the conditions of urban life.

Sources

• Empirical sociology (Wikipedia)
• Chicago school (sociology) (Wikipedia))
• Albion Woodbury Small (Wikipedia)
• Life and Labour of the People in London (Wikipedia)
• Paperspast: 1894 report on Booth's poverty figures (National Library of New Zealand)
• Survey Maps: Mapping Economic, Social and Racial Inequities (University of Toronto Fisher Library)
• The Philadelphia Negro, Encyclopedia of Greater Philadelphia
• W. E. B. Du Bois and "The Philadelphia Negro", Roper Center, Cornell University
• The Polish Peasant in Europe and America (Wikipedia)
• Gettys (1959), review of The Polish Peasant, Mead Project, Brock University
• Middletown studies (Wikipedia)
• Robert and Helen Lynd, Encyclopaedia Britannica
• Otto Neurath (Wikipedia)
• Marienthal: The Unemployed, Haus der Geschichte Österreich
• The Marienthal study, Archive for the History of Sociology in Austria, University of Graz
• Suicide (Durkheim book) (Wikipedia))
• Levitt and List (2009), Was There Really a Hawthorne Effect at the Hawthorne Plant? NBER Working Paper 15016
• The Sociological Imagination (Wikipedia)
• Positivism dispute (Wikipedia)
• Tearoom Trade (Wikipedia)
• Belmont Report (Wikipedia)
• General Social Survey (Wikipedia)
• The General Social Survey, NORC at the University of Chicago
• European Social Survey (ICPSR 24190)
• Response rates in telephone surveys have resumed their decline, Pew Research Center (2019)
• Lazer et al. (2009), Computational Social Science, Science (Gary King's site)