Well first of all I am probably going to choose something I see as a problem that needs solving. Let us say car crime.
Do I start from a theory or do I start by observing the problem, gathering data etc.?
Where do I start in designing my research.
Leads on to INDUCTION vs DEDUCTION.
INDUCTION assumes science/social science starts from observation. The starting point is observation from which we make connections or correlations. We come to understand the world through our experience of it. This is sometimes called primitive empiricism.
Inductivists argue that we look at the evidence and move from that to a theory.
DEDUCTION Popper and others argue that(social) science advances through theory testing.
We start from theory and make observations to test theory. IF the observations are in line with the theory then the theory remains and is bolstered. (Popper uses the term verisimilitude or truth likeness).
Deductivists argue that we can’t observe the world without a theory guiding our observation. So how can I study car crime unless I have a prior idea of what is relevant to car crime? We all have thoeries or explanations which govern our observation of the world, we do not simply
go in and look at a situation and then develop theory.
The deductive approach is often seen as being more scientific as it mirrors the experimental approach in natural science. We have a theory which makes predictions about the world. We then look for observations which counter that - we are trying to falsify a theory. And if we fail to do so it survives.
Stages Of Research Process
DEDUCTION INDUCTION
Theory Observation/Data Collection
Hypothesis Analysis
Observation/Data Collection Description
Data Analysis Formulation Of Theory
Findings
Feeds Back In To Theory
In practice deductive and inductive reserarch tend to have different roles.
Inductive research tends to be used where there is no well established body of theory. Deductive or theory testing is more commonly linked to areas where there are established theory which can be tested against empirical evidence.
They are linked in to another commonly made ddistinction in social research - that between qualitative and quantitatvie research. Qualitative research is typically that which relies on interviewing and observational methods. It seeks in depth data emphasising the perspective of the subjects of the research.
Quantitative research is that which measures the incidence of different types of behaviour, attitudes or characteristics. It looks for correlations between them - for example class and voting behaviour. Because it is working with measurable things it tends to use statistical techniques in looking for relationships.
. Qualitative and Quantitative
The distinction between qualitative and quantitative research is commonly made.
Quantitative research is sometimes referred to as statistical or sometimes (somewhat sloppily) as empirical research.
Qualitative research is sometimes referred to as the sociological approach to research.
Elsewhere the distinction is often made between certain research methods as either being Qualitative or Quantitative.
For example Interviewing, Participant Observation and Documentary Analysis are often referred to as Qualitative research methods.
The use of Surveys and Statistical data is referred to as Quantitative.
These distinctions do have a fair degree of truth to them. However as we will see later they do not hold true in all cases. What are commonly thought of as used in purely Qualitative research can be part of Quantitative research programmes and vice versa.
In this lecture we are going to look at this distinction between qualitative and quantitative research. After looking at brief definitions of each we are going to look again at what the goals of social research are. This will allow us to reflect on how both qualitative and quantitative research strategies can be employed to achieve those goals.
Definitions
“Qualitative Research is a basic strategy of social research that usually involves in-depth examination of a relatively small number of cases. Cases are examined intensively with techniques designed to facilitate the clarification of theoretical concepts and empirical categories.” (CC Ragin, 1994, p190)
“Quantitative Research is a basic strategy of social research that usually involves analysis of patterns of covariation across a number of cases. This approach focuses on variables and relationships among variables in an effort to identify general patterns of covariation.” (CC Ragin, 1994, p190)
Each can make a contribution to the study of society.
The best way of approaching their relative strengths and weaknesses is through looking at what are the goals of social research.
7 Goals of Social Research (from CC Ragin)
1 - Identifying General Patterns and Relationships
Thinking back to last week we discussed how one of the key features science is the identification of General Patterns and Relationships. Those that see social research as a true science often see this as the Primary Goal of research.
Knowledge of general patterns and causes of social phenomena are often valued because they can help us address social problems or act to guide policy.
For example, if research were to suggest that poor attendance at school were a cause of crime then that would give us a possible approach to tackling crime.
Social researchers may believe that discovering general relationships is best done through the examination of many cases. THe reasoning being that certain cases may be atypical but that by looking at many cases we can identify generalities.
2 - Testing and Refining Theories
General patterns are especially relevant to social theory. The implications of theories are tested and refined or discarded. Typically this is done according to a general plan of scientific method.
Hypotheses are derived from theories and their implications and then tested with data that bear directly on the hypotheses.
In theory we abandon a theory if it is falsified by an observation.
In practice a single piece of counter evidence or a single unsupported hypothesis usually does not result in abandoning a theory. Rather unsupported theories gradually fade from current thinking or are refined.
3 - Making Predictions
Social researchers use accumulated social scientific knowledge to make predictions about the future and other novel situations.
Predictions can be made from two types of knowledge.
Firstly the use of Historical Knowledge.
e.g. The Stock Market Crash of 1920 and ensuing Great Depression has motivated our economic and political elites to moderate the violent swings of market oriented economic life.
Secondly the use of General Patterns.
For example we know that certain types of criminal activity - drug dealing - increase when legitimate economic opportunities decrease. We can use this knowledge to predict future crime rates based on economic predictions.
Predicting rates - crime, unemployment etc - is much easier than predicitng specific events such as the fall of the Berlin Wall and suchlike. At best social scientists are making broad predictions about trends.
4- Interpreting Culturally or Historically Significant Phenomena.
Certain atypical events are interesting because of the role they played in establishing the current situation. For example the US Civil War has an impact on current power relations. There may be competition between different interpretations of events to become the accepted interpretation.
In fact often there is no single accepted "correct" interpretation of an event and different analyses are seen to bring different understandings.
Research which seeks to explain particular events often addresses issues related to the consciousness of the actors involved - whereas research which sseks to establish generalisations often does not.
In other words the intentions of particular people involved in events are often addressed in event specific research.
5 - Exploring Diversity
The goal of exploring and comprehending social diversity surrounding us.
This is related to the search for general patterns but instead seeks to understand the deviations from those patterns.
Ragin argues this goal of social science is about the study of unusual social practices and phenomena.
6 - Giving Voice
Research which seeks to tell the story of, raise understanding of or give a voice to a group in society.
The researcher often has to unlearn or disregard established theories to present the subjects' interpretation of the world.
There is a debate as to whether such research is simply a type of advocacy for a given group or way of life. Even if it is it may be valid in offering an alternative conception to the mainstream.
7- Advancing New Theories
Theory testing (goal 2) is usually primarily deductive.
By contrast Theory advancing research is usually described as inductive.
The classic conception of such research is of in depth case studies leading to an understanding of a social phenomenon from which new theories, ideas or concepts are generated.
So whereas for deductive research the theory comes first and observation is collected to test the theory.
For inductive theory advancing research - first we collect data or investigate a subject and formulate theory from that.
It is often not possible or desirable to address all or most of these goals tegether in one piece of research.
For example the different stage of data collection creates difficulties if we attempt to both test and develop theory in the same piece of research.
Relation of 7 Goals to Qualitative / Quantitative distinction.
So the goals of qualitative research are primarily to advance new theory, interpret the significance of individual events and giving voice to particular groups.
In contrast Quantitative research is primarily about Testing theory, identifying broad patterns and making predictions.
So the choice of research strategy will depend at least in part on what we want to achieve.
For example take two studies which concerned delinquency amongst the young.
Hirschi wanted to test established theories of delinquency. In his 1969 study he adopted a Quantitative research strategy whereby he surveyed 5,000 teenagers.
In contrast Patrick in his study of delinquency used covert particpant observation to research a 1960s Glasgow gang in A Glasgow Gang Observed. Such a study has the capacity to develop theory of delinquency and to provide a deeper understanding of the gang phenomenon in Glasgow. However it is less able to test general theories of delinquency than a study such as Hirschi's.
Quantitative researchers are interested in how variables covary across cases. Therefore they seek a limited amount of information about many cases.
Qualitative research by contrast seeks a lot of information about a limited number of cases.
Again this is demonstrated in the examples of Hirschi's and Patrick's studies.
What Are Qualitative/Quantitative Methods?
Certain methods are generally said to be Qualitative or Quantitative.
However those methods commonly classed as qualitative can be used in quantitative research and vice versa. For example content analysis of documents -
e.g. Ian Budge measuring what % of manifestos devoted to certain issues.
E.g Systematic observation - counting of instances of relevant behaviour which are then analysed statistically.
THerefore research methjods can't be classified as simply qualitative or quantitative - it depends how we use them to collect the data and how we analyse the data we have collected.
How Else Do We Choose Methods?
So far we have discussed how our goals influence our choice of research strategy. There are a number of other factors which influence the choice.
Cost - Quite simply certain research strategies are more expensive than others. Quantitative research can be quite cost effective if we use available statistics.
Much social research is done using the Census, General attitude Survey, publications like Eurobarometer and other official sources. Furthermore the ESRC data archive at Essex University holds a mass of data from previous research projects which can be reused in an original way. The reuse of qualitative data such as interviews is possible but is unusual - partly because much less is available in the public domain.
Qualitative research is expensive because it is time consuming. For example a one hour interview which is taped can take 3 hours to transcribe before analysis even begins.
Reactivity - Certain research strategies suffer from the problem of reactivity. That the knowledge that they are being researched alters the behaviour of the subject. This is really a problem for qualitative research strategies such as interviewing and overt particpant observation. Again going back to the Patrick study only one member of the gang he joined knew he was in fact a sociologist doing research. He kept this quiet because it is unlikely the gang members would have acted naturally in his presence had they known the truth.
Suitability - Certain subjects may simply be unsuitable for study through certain methods. FOr example sensitive subjects like bereavement are probably no tsuitable for impersonal approaches like surveys. You wouldn't send a questionnaire to people recently bereaved - insensitive.
Similarly some important or self importatn people might expect a personal interview and be insulted to receive a questionnaire in teh post. Others may prefer to fill in a questionnaire as it may be quick and impersonal.
Availability - Some subjects are difficulkt to get data on using certain research methods. FOr example a swe discussed - criminal behaviour may be difficult to research by interviews as people may be unwilling to speak and incriminate themselves. Researchers may have difficulty gaining access to figures like top politicians and so may have to supplement research with available documentary evidence.
Problems of interview access are greatest when the subject is dead! A serious point - read biographies, personal papers etc.
Strengths/Problems of QUalitative Research
Qualitative Research
Contextual - It sets explanation in context.
However the down side of this is that explanation is IDEOGRAPHIC rather than NOMOTHETIC.
In other words any explanation is linked to a specific place and time and this prevents generalisation in the same way as Quantitative research which is generally NOMOTHETIC - law like.
Flexible - Because of the way theory can emerg during the process it provides a flexibility to the research. It allows that we may have missed something very relevant to the subject that we are researching and for that to emerge during the research.
QUantitative research requires much more specification of what is relevant at the beginning of the process. We need to know what theories to test, what data to collect, what questions to ask on a survey and so on.
Interpretaion of meaning. Because qualitative research is concerned with the meaning people attach to action it creates a difficulty of how do we know our interpretation of the meaning of their behaviour is correct. Weber argues that on way is through the concept of VERSTEHEN - understanding. That because we are human we can empathise with other people. This is an interesting approach but seems rather haphazard. However often we do rely on these kind of judgements.
The differences in the Qualitative and Quantitative are summarised in the table below.
So to sum up whether we choose qualitative or quantitative research should properly depend on objective judgements on criteria based on how we best advance knowledge. Often this will mean that qualitative research is used as preparatory research to generate theories and understanding of a field. When this is established quantitatve research can be used to test those theories is a more rigorous way.
However as we discussed at the beginning this isn’t always the case. For example , Okely (1987) wanted to study gypsies in the UK. It was hoped that the research would inform policy towards the gypsy community.
Okely judged that qualitative research was best suited to this reserarch - given that relatively little was known about the gypsy community, their beliefs, culture and lifestyle. However, pressure was brought to bear on Okely by the research funders on the grounds that only so called facts - in other words measurable phenomena - behaviour, beliefs, characteristics - would influence policy makers.