Research methods Flashcards
Positivism; key facts about their research?
- identify patterns and trends in data
- make comparisons of data over time
- finding correlations to identify cause and effect relationships between variables
- value free and objective research
- quantitative data collection (numerical data is favoured)
- operationalise key variables and concepts
- experiments (laboratory and field) are repeatable and therefore results can be checked and verified
- Durkheim follows the Positivist approach
Positivism; disadvantages?
- social reality cannot be studied in the same way as the natural world (natural sciences) because in the social world people interpret situations in different ways.
- they ignore human free will and rely on laboratory experiments as a research method
- ignore the subjective experience of individuals focusing on objective approach (ignore interpretations and meanings that people place on their actions)
- statistical data lacks verstehen (depth, meaning and reasoning)
Interpretivism; key facts
- qualitative data that’s rich, detailed, high quality and descriptive that shows meanings and motives
- reject positivism and natural sciences because we’re dealing with humans in sociology and so meaning is attached to how they act
- important to gain a rapport with the people being researched as this increases validity
- gives us good verstehen (empathetic understanding)
- aim to be reflective to examine if the participants’ behaviour changes during research
- participant observation, unstructured interviews and personal documents etc
Interpretivism; disadvantages?
- usually focuses on small samples and they might not be representative of a wider group or population in general.
- difficult to replicate because much of the research is the product of the unique rapport that the researcher has built up with members of the group. Another sociologist would produce a different set of data therefore reliability cannot be guaranteed (unreliable)
- research may be biased because it’s not value free
10 Main stages of the research process (area of interest to conclusion)
- Area of interest (choice of topic)
- Background research (helps to formulate your research question)
- Hypothesis
- Choice of research method (qualitative or quantitative)
- Detailed planning of your research (eg operationalise key concepts)
- Pilot study (testing out questions)
- Revise your methods if necessary (are they valid and reliable?)
- Carry out the research
- Analyse the results
- Conclusions
Define validity?
Whether the research findings give a true picture of what is being studied
Define reliability?
A reliable measure would produce the same results on different occasions
Define representativeness?
The people who take part in the research should have characteristics that are typical of the wider population being studied. They should come from the same or similar socioeconomic backgrounds and reflect the same ratio of gender, age and ethnicity.
Define generalisability?
Means that the sociologist can safely conclude that what is true of the sample that was was actually studied is probably true of the wider population to which the sample belongs.
What’s a longitudinal study?
It’s carried out over a period of weeks/months/years in order to study changes or developments over an extended period of time, it can also be used to identify casual relationships and correlations between variables eg health and class
- the sample must be carefully considered, should remain relatively large and should aim to prevent participants from dropping out
- how will it be funded
- ethical issues need to be considered
Longitudinal Studies; What’s a panel study?
The data is collected from a sample selected from sampling frames such as the postcode address file. The sample unit may be individuals, households or organisations.
Longitudinal Studies; What’s a cohort study?
They aim to study people who have the same social characteristics eg age. A whole group of people might be studied eg a class at school or a sample or the class might be studied.
Longitudinal Studies; methods?
Interviews
Observations
Ethnography
Longitudinal Studies; Advantages? Disadvantages?
- identify trends and patterns and correlations over time
- verstehen
- good rapport with the participants, more comfortable and open, results might be more valid
- can be expensive
- time consuming
- respondents may drop out which results in a smaller sample and this might not be representative of the population / group
- relationships built may mean the results are not reliable because they’re biased
Primary data ?
Eg Interviews, Observations, Questionnaires
Which enable new data collection that did not exist before the research began
Positivism - use quantitative methods eg a questionnaire or interview with a scaling system. Quantitative data eg stats/facts/figures are More reliable but less valid because they lack depth.
Interpretivism - qualitative methods eg unstructured interviews and observations. More valid because it shows meanings and motives but less reliable because replication of the same data/results might not be possible
Choosing a primary data research method?
- reliable = same/similar results produced by repeating method
- valid research = provides a true picture of what’s being studied or measured
- practicality;
Quantitative > qualitative because it’s more practical and less time consuming and allow you to study a large sample of people
Primary data; lab experiments?
Experiments conducted in a lab tend to use scientific equipment to measure the variables and concepts.
Advantages = variables can be controlled and experiments tend to have high control. Reliable as used in natural sciences. BUT the settings are generally artificial and we can’t study humans using that, it doesn’t give a true representation (less validity).
Primary data; field experiments?
Conducted in a natural setting/situation eg a classroom. Behaviour can be studied with some form of control.
BUT less control = results can be inaccurate. Hawthorne effect, people can change their behaviour because they’re being studied so giving inaccurate results,
Primary data; case studies?
These examine one particular case or instance eg a life history of an individual.
Advantages - Can be used to prove or disprove a hypothesis. Give new info or insight into a topic. Can become a pilot study.
BUT can be limited, unrepresentative of the general population.
Secondary data key points?
Official statistics is the best example. Form of secondary data which collects data in a quantitative manner that would be favoured by Positivist sociologists but rejected by interpretivists.
Secondary data; Official statistics key points?
The best source come from the census which is undertaken every 10 years by every household in the UK.
But can also be collected from
- registration
- official surveys (ONS responsible for all UK stats)
They tend to be reliable but not very valid
Governments may however manipulate statistics to their advantage which is called political bias therefore reporting them in a way making them not very valid. Eg governments might hide some data, released when there’s other news in the media dominating BUT if this is the only data available then sociologists have the rely on this
Secondary data; Advantages of official statistics?
- well planned, organised, detailed, official statistics should be quite reliable.
- representative samples are used to collect official statistics and these samples are very large (much larger than the sociologist could conduct alone due to lack of funding) eg census = whole of uk
- some research is conducted regularly and therefore has high temporal validity because the research is up to date and current
- published stats are usually readily available for everyone to access and use and they cost nothing or little to access eg through the internet
- patterns and trends of data are usually very easy to identify and recognise and this can be very useful when analysing large amounts of official statistics
Secondary data; Disadvantages of official statistics?
might not be fit for purpose if they’re not measuring what they intend to do eg annual crime stats by the home office don’t provide an accurate representation of crime due to unreported and unrecorded crime not part of this.
The government compiles and collects official stats so they might be biased in their political outlook and therefore not neutral.
Not very valid as The figures do not express reasons for trends/patterns and they lack meaning. Everything is reduced to a number and means that official statistics can be low in validity.
Secondary data ; non official stats?
Stats collected by independent companies, researches or academic institutions. QUANTITATIVE
+
Objective quantitative data
Quite accessible
-
Can’t guarantee reliability
Can be expensive
Secondary data; family histories?
Recorded histories left by those who have experienced past events. Eg sound recordings, film or old photographs which can provide insight into the past. QUALITATIVE
+
Rich and detailed
Better insight into culture, value and norms
-
Difficult to compare
May be biased to make that family look good etc
Time consuming to analyse
Secondary data; previous sociological research ?
We can examine the new research topic, aim and results and then decide if we would like to replicate the study or extend the research topic further into a new area.
QUALITATIVE / QUANTITATIVE
+
Saves time and money
Gives more data and allows for comparisons
-
Might be biased
Could be out dated
Could be inaccurate
Assessing secondary sources
Authenticity - how genuine the documents are.
Authorship - who wrote the document
Credibility - amount of distortion
Availability - some documents aren’t available
Define operationalise
Define a key word or variable and state exactly how we’re going to measure them
Preferred by Positivists
To aid reliability (enable replication by other researchers
who can use the same operationalisation)
To aid objectivity
To aid validity (measuring what you are setting out to
measure)
Eg educational disadvantage can be operationalised by using proportion of students on free school meals
What are social problems?
Aspects of social life that cause misery and anxiety to both private individuals and society in general. Eg crime, child abuse, domestic violence and suicide.
What are sociological problems?
A successful society involves a number of social institutions effectively carrying out a range of social processes and functions. Sociologists are interested in how societies and social institutions work and how individuals interact and interpret one another’s behaviour. So any aspect of social life normal or deviant is a sociological problem. Eg sociologists are interested as to why divorce is a sociological problem because it can have negative effects for individuals and society but equally marriage is a sociological problem because why is it important and what makes it good or bad etc. Overall research is to increase knowledge of how society works and why people behave the way they do.
Social Policy
Laws and Regulations
Eg as of April 2019 in Scotland domestic abuse also includes psychological abuse
What’s the relationships between social problems / sociological problem and social policy?
- Some sociological research simply aims to be descriptive. It’s purpose is to describe a particular situation that or set facts.
- Some sociological research is explanatory. It sets out to explain a particular social phenomenon such as poverty.
- Some sociological research is evaluative. It sets out to monitor and assess the effect of a social policy.
Practical factors that affect choice of sampling process?
Cost
- how much funding the research has managed to secure.
- generous = large interviewing team /national survey
- not much = cheaper options eg postal questionnaires or secondary sources (cheap and reliable sources of data)
Time available
- dependent on funding eg lots = longitudinal study possible
- lacks time = a small scale study and maybe even secondary rather than carrying out first hand research
Subject matter
- how comfortable people are to discuss the topic eg sensitive areas = people might not want to participate so maybe anonymous questionnaires might be better
Theoretical factors that affect the choice of sampling process?
Influenced by positivism / interpretivism approach
Respondent validation and interpretation of data
The sociologists interpretation of an event or data is checked with those who took part in the event. Feedback is obtained from the participants about the accuracy of the data and about whether the researcher has fairly interpreted their behaviour.
Mainly used for interpretivist research
It’s to reduce the possibility that the researcher applies their interpretation to the data BIAS
Define sampling unit and sampling frame?
These are members of the population that you wish to study see in terms of number and characteristics eg; 40 male dentists. Some sample units might be difficult to obtain.
Once the research population has been identified eg mothers aged 20-30 in the UK; a sampling frame is used which consists of a list of members from the population that you wish to study. The source for this could be the electoral roll, telephone directories etc.
Gate Keeper?
Controls the access to the sample. Sometimes it’s easier than others.
Eg a school = headteacher
Eg a gang = Gang leader (more difficult to find)
Sampling process?
Precised definition eg focusing on the poor, what characteristics defines someone as poor
Find access to the research or target population
Sample has to be chosen. This has to be a representative sample in order to ensure generalisability (generalise that the findings apply to the wider research population)
Random sampling?
Every member of the sampling frame has an equal chance of selection eg picking names out of a hat
+ limits bias within the selection process as samples are chosen by chance
- the sample might end up being non representative of the population eg only males are selected
Systematic sampling?
Randomly choosing a number between one and ten eg 7, and then picking out every tenth number from that number eg 7,17,27 etc until you reach the required number eg 30.
+ limits bias within the selection process as samples are chosen by chance
- the sample might end up being non representative of the population eg only males are selected
Stratified sample?
Dividing the research population into a number of different strata in terms of characteristics. A sample is then selected that reflects these characteristics eg 51% female and 49% male.
+ it will be representative (even proportionally) of the research population.
- it’s time consuming to select the sample and you need a complete list of the population which might be difficult to obtain.
Snowball sampling?
Someone willing to participate in the research is found eg gang member. The participant therefore is then asked to find another person to take part. A network is built up and the sample grows.
+ gives access to a group of people where there’s no sampling frame available or they engage in deviant / illegal activities, normally carried out in isolation or secret. And this sample can grow as large as the researcher wanted.
- people taking part might not be representative of the research population and the researcher has little control over the sampling method.
Volunteer sampling?
Researchers try to attract volunteers by using leaflets and posters etc. People can read the information and can decide whether or not they want to volunteer their time.
+
Minimal time/effort is required because the sample comes to you. And those who volunteer themselves are more willing and therefore provide more insightful data
-
People who take part might not be representative of the population and you might not even get a large enough amount of people to conduct the research ie low participation rates
Opportunity (convenience) sampling?
Selecting the participants in situations / opportunities where the research population is likely to be found. Sociologists might have easy access to this group and have contact with them. Therefore it’s convenient to ask them and taking whatever you can get.
+
It’s convenient and easy to find participants and build a good rapport with them
-
Can lead to an under representation or over representation of certain groups within the sample and it’s not ideal because your limited to an opportunity that might not respond as well as you’d hoped it too
Purposive data sampling
Used by researchers who know what they’re looking for and go and find that specific sample. Eg; wishing to find 25-50 year old doctors so going directly to hospitals.
+
Easy to directly find participants fitting of the criteria and prevents irrelevant people from being part of this sample.
-
Prone to researcher bias and because they’re chosen from the same place (eg doctors from one hospital), it might not be representative of the whole research population
Quota sampling ?
Researcher looks for the right number of people until the quota is filled eg; finding 20 females who work full time aged between 30-45. The researcher uses the first 20 ladies who match the criteria and agree to participate.
+
Prevents irrelevant people from being part of the sample. And not dependent on a sampling frame.
-
Danger of use researchers only question those who look “suitable” and “co-operative”. So it’s difficult to say whether a sample is truly representative of the research population.
What are Pilot studies?
A pilot study is a small scale preliminary study that sociologists might conduct before they do their main research. It’s like a test run using a small sample of participants and is used to see whether the research will actually work. It can be used to decide whether the study is valid, whether any changes need to be made in the design etc. They’re commonly used when the research involves the collection of quantitative data.
Evaluation of pilot studies?
+
Problems and issues can be rectified
You can evaluate the practicalities eg timings and method and whether the questions were understood
You can find out if the study’s worthwhile and practical and is the data produced the kind that was wanted.
-
Time consuming
Expensive
Cannot use the same participants for pilot and main study so you have to find two samples.
Ethics?
It’s important that sociologists follow the guidelines provided by the British Sociological Association BSA. Sociologists should try to follow these guidelines at all times.
However BULMER argued that sometimes the researcher will have to compromise and make sacrifices in order to protect themselves and the participants in research.
Principles of ethical research by Bulmer?
Informed consent (choice whether to participate and full info of the study)
Respect for privacy (shouldn’t invade privacy or make them feel uncomfortable)
Protection from harm (no harm physically, socially, psychologically)
Safeguarding the confidentiality of data (anonymous)
Avoid deception (deceit and lies should be avoided however if it’s necessary as it compromises the validity of data then debriefing is required after)
Awareness of the consequences of publication (participants aware)
Right to withdraw (they can refuse whenever and don’t need a reason to)
Ethical issues in quantitative research ?
Before the research begins, participants need to be informed about the nature and purpose of the study, how their data will be used, how anonymity will be respected and the confidentiality of their answers
It’s important to build a rapport with the participant before an interview takes place. This will ensure that the participant feels confident that they can refuse to answer any questions that they find uncomfortable. Sometimes questions that appear neutral can still cause upset, so be aware of this.
Large scale questionnaires generate huge amounts of data which must be kept secure in line with the data protection act. Many questionnaires do not require the participant to give their name, address or date of birth (which ensures confidentiality)
Ethical issues in qualitative research?
Sometimes it might not be possible to gain fully informed consent from a person who is being observed but there’s ways around this eg asking them after the observation.
Eg covert research like Glasgow gangs by James Patrick. They’re breaking the ethical guidelines as no fully informed consent gained, the participants are being deceived and privacy is invaded.
But it’s justifiable because you need deception otherwise demand characteristics will show (hawthorne effect) and u won’t get valid results. But also in order to study sensitive topics eg gangs.
Questionnaires?
- Mainly used to collect quantitative data especially if closed questions are asked and this is favoured by positivist sociologists. - consist of questions compiled by researchers that are then distributed and need to be self completed by participants (usually not in the prescience of the researcher)
- are suitable when the research topic requires numerical answers but not so suitable for studying sensitive topics eg domestic violence.
- should reach a suitable sample of respondents and should produce a high return rate in order to gain valid results
Constructing a questionnaire?
A poorly designed questionnaire could make the results invalid and inaccurate so it’s important to spend time and effort on careful construction.
- clear and should be the same for all participants (increases reliability)
- should be worded clearly so it’s understood the same (increases validity)
- where and when to return the questionnaire needs to be stated
- purpose should be explained
- confidentiality should be ensured
- shouldn’t be too long or short
- questions should be biased (not leading or biased)
- shouldn’t be offensive
- open require free response - enough space
- closed require yes/no or tick boxes and consider the bands don’t overlap
Structured questionnaires facts?
Consists of closed questions that require fixed answers
Favoured by positivist sociologists where respondents can answer the questionnaire quickly by ticking boxes or selecting answers etc
Data will consist of figures,percentages,statistics (quantitative data)
It’s easy to analyse
Questions usually consist of key variables that must be operationalised eg key variables are defined and measured
Reliable because if handed out again you’d get the same results
However interpretivists would criticise and say its low validity because real life is too complex to categorise Via closed questions and fixed responses.
Closed questions should follow 3 main principles?
- Exhaustiveness
Categorises should account for every possible response, and to cover this issue sometimes the categories of other or none of the above are used. - Exclusiveness
Respondents should be able to select only one response from the ones available in order to make analysis of results easier (band categories) - Balancing
When respondents need to place attitudes on a scale, there should be equal amounts of positive and negative categories in order not to distort answers.
Unstructured questionnaires facts
Consists of mainly open questions that require respondents to answer the questions themselves in their own words
Favoured by interpretivist sociologists where high quality answers are given that might be in depth - looks at themes, meaning and motives
Data will be in depth and high quality (qualitative)
Data can be difficult to analyse because each participant might have given very different answers and different length answers too. There might be little similarity of themes / content in the answers given
The questions can consist of key variables that must be operationalised
Semi structured questionnaire facts?
Open and closed questions
Favoured by realist sociologists
Quantitative and qualitative data is collected
Variations of this type eg;
- a self report = lists a number of items or activities and asks respondents to tick those they have experienced
- attitudinal questionnaires = ask respondents on a scale of 1-5 whether they subscribe to a particular point of view
Different types of questionnaires?
Self completion by post;
- include a covering letter explaining the aim and purpose and give a stamped addressed envelope so it can be easily returned.
Internet based surveys
- sending respondents to a direct website where they can fill it in eg survey monkey
Hand delivered
- researcher delivers it by hand to the designated company/person/household
Face to face on the street
- stopping people on the street and getting them to fill it out straight away
Advantages of questionnaires?
- Can reach a large and consequently a more representative sample as it can easily be handed out or posted to lots of people.
- If anonymous, they’re useful for research that intends to ask embarrassing or sensitive questions such as sexual behaviour.
- Researcher has minimum contact with the respondent so they can’t directly influence the results by being present (Hawthorne effect)
- Positivists believe that if the questions are neutral and objective then another researcher should be able to repeat and get similar results therefore it’s high in reliability.
- Postal questionnaires can be used especially if the research population is geographically dispersed across the country or if results are required from different regions for comparitave purposes.
Disadvantages of questionnaires?
- Many people cannot be bothered to reply especially postal ones, therefore you can suffer from low response rates to even no responses. Compromises level of generalisability if the sample size is smaller.
- Interpretivists criticise them for producing data low in validity because they argue that real life is too complex to categorise in closed questions and fixed responses.
- Interpretivists suggest that there’s a danger with respondents interpreting the question in a different way that was intended and then the researcher can’t clear up any misunderstandings therefore they’re not VALID
- Interpretivists argue that closed questionnaires with tick box responses suffer from the imposition problem. This means that the sociologist has limited responses to measure what they think is important which might force respondents to think in the same way. Which might frustrate respondents and they may leave it incomplete
- People may lie or misrepresent themselves to give an impression of a normal citizen eg lying about mental health.
Structured interviews? (Positivists)
Researcher and the interviewee will have interaction and a rapport will be developed between them. The researcher can also note down any non verbal aspects of behaviour eg body language. They should ask questions in a clear standardised way in a good order that makes sense but also record the answers carefully while maintaining a rapport with the interviewee.
Disadvantages of structured interviews?
- The researchers characteristics eg age gender ethnicity, may have an effect on the interviewees answers which will affect the validity of the results - this is called the interviewer effect
- Social desirability bias means that the interviewee might lie in order to portray themselves in a good light eg how much alcohol they drink which once again will affect the validity of the results.
- Feminist sociologists would criticise structured interviews because it creates a hierarchy between a powerful researcher and an interviewee with no control. They’d prefer semi structured or unstructured because it breaks down the barriers between the two.
- Lacks freedom so both the interviewee and interviewer cannot explore any other ideas or areas relating to the subject.
- More costly in terms of time and resources than a postal questionnaire which can be filled out in the interviewees own time.
Advantages of structured interviews?
- They’re quick and easy to administer so a large sample size can be potentially be used.
- Researchers can easily be trained to conduct these interviews because it’s just following a standardised set of instructions and this is usually inexpensive.
- Positivists would prefer structured interviews because they consist of closed questions so the answers can be easily quantified as the answers are straight forward.
- Researchers are trained to ask each question in the same way and not to get too familiar with the interviewee, this reduces interviewer bias but also ensures consistency and reliability of the data collected.
Semi structured interviews ?
Many sociologist interviews are a mix of structured (closed questions to get facts) and unstructured (some open questions to get extra detail)
Semi structured interviews advantages?
Open and closed questions, gain quantitative and qualitative data which is very useful (positivist and interpretivist data)
Reliability can be enhanced if questions are planned in advance so the interview can be repeated and results can be checked for consistency.
They can be quick if lots of closed questions are asked so a larger population size can be targeted and a great deal of data can be gained.
Semi structured interviews disadvantages?
Validity may be limited to the few open questions used however still more than structured
But if open questions are asked then the reliability of the interview can be limited
Can be time consuming if there are many open ended questions and therefore the sample size might be smaller than that of structured interviews.
Unstructured interviews?
They’re essentially guided conversations where the talk is informal but the interviewer plays an active role in order to keep to the subject of the research.
Flexible as follow up questions etc can be asked to probe responses and investigate motives and feelings in a way that structured interviews don’t allow.
Interpretivists argue that it’s ethnographic because it’s carried out in a comfortable natural setting and theyre in depth and carried out over hours rather than minutes.
Unstructured interviews advantages?
Interpretivists would advocate this method because verstehen and insight can be gained from the rich and vivid data in the form of extensive quotations - data is therefore highly valid.
Good rapport can be established between the researcher and interviewee therefore they can research sensitive topics that involve empathy and trust OR the interviewee will open up more and give answers they wouldn’t put down in a questionnaire etc so the results may be extremely VALID.
It’s more flexible as the interviewer can clarify questions for the interviewee, they can ask follow up questions and are not restricted to a list of fixed questions only as they’re controlling the direction of the interview.
Unstructured interviews disadvantages?
Positivists would criticise this approach because the qualitative data is hard to analyse because of the volume of material in the respondents own words, it’s impossible to quantify this and turn it into graphs etc.
Very low reliability as it cannot be recreated, much of the content is due to the rapport built between the researcher and researchee and so cannot be recreated to produce the same results.
Expensive to carry out because training needs to be more thorough and specialised and focused on interpersonal skills, but also time consuming because they take several hours to complete. Overall this means that a smaller sample might be used in comparison to structured interviewers, which means that it’s less representative of the research population as a result so it’s difficult to generalise to the wider community.
Interviewer bias and social desirability bias and interviewer effect can effect results and lower validity.
Content analysis?
This is used to analyse the content of a text or source. It’s a way of quantifying data in the form of pictures or writing (quantitative data analysis). It aims to quantify content in terms of pre determined categories in a systematic and replicable manner. Content analysis has the potential to uncover hidden aspects of sources/text and bring them to the surface.
Eg fawbert studied mass media by looking at how many times the word hoodie appeared in newspapers and wanted to link this to moral panics.
Positivists favour it because it focuses on quantitative data and deals systematically with contents of documents etc.
Content analysis stages?
Choose an appropriate set of texts, songs, pics
Break them down into smaller units that can be easily analysed eg a time frame / paragraph etc
Decide on the relevant categories and their headings
Analyse the units according to the categories
Count how often each category occurs
Analyse the findings in terms of frequency and any patterns
Content analysis advantages?
It’s a cheap method as you only need to purchase the magazines/newspapers/watch tv etc
It’s a comparative method as it allows you to compare media reports and content over time (can be longitudinal)
It’s a reliable method because sociologists can repeat and cross check the results by looking at the same data and using the same categories etc.
Content analysis disadvantages?
It’s very time consuming because media needs to be checked over a long period of time
Subjective method as the categories largely depend on what the sociologist thinks is important
Sociologists who have used this method have been accused of analysing text out of context
The Media products themselves may only tell us about the personal and political beliefs of those who produce them eg the prejudices of journalists.
Ethnography?
Is the study of the way of life of a group of people, their culture and the structure of their society. Researchers aim to “walk in their shoes”. Researcher participating in or observing their daily activities. Other activities eg unstructured interviews, may be used to sketch out a fuller picture of the groups behaviour
Ethnography Advantages?
Preferred by interpretivist sociologists who argue that eg unstructured interviews, allow researchers to gain full verstehen (being able to empathise with and or think like the people who are being studied)
And get access to the lived experience of particular social groups. So the closeness to the research subjects has led interpretivists to claim that ethnography produced the most valid and authentic type of qualitative data of any type of social research method.
Disadvantages of ethnography?
Could be difficult to gain access to rage participants and there be ethical issues eg invasion of privacy and deception
Time consuming - it can be longitudinal and requires a lot of effort for the researcher
Interpretivists and Observations?
Looking at the topic that’s being studied and recording data. Interpretivists sociologists would analyse meanings from observations which can be very subjective since the sociologist chooses what behaviour to observe and make notes on (qualitative data).
Positivists and observations?
Positivist sociologists are more systematic in their approach and might observe data using charts (quantitative data). Positivists would see observations as an essential research method, especially when conducting experiments where behaviour needs to be observed. Positivists are less likely to become involved with participant observation (but would use non participant observation instead).
Participant observation ?
Sociologists can join a group and observe a group of people. Sociologists can participate in their daily activities and observe what people say and do. Griffin (1960) a white journalist dyed his skin black to see what it was like to live as a black man in the southern states of the USA in the late 1950s. Interpretivists would advocate the use of participant observation.
Can be covert or overt. Type of ethnography. Most common type of overstating to ‘get inside their heads’
Gaining entry to participant observation?
Participant observation can only work if the researcher gains entry into the group and can be accepted by the group members. Many groups do not want to be studied especially if they’re deviant or criminal. However it is possible to enter closed groups through a gatekeeper (particularly in overt)
Conducting research (participant observation)?
Participant observation involves looking and listening and the researcher should be able to ‘go with the flow’ rather than forcing the pace and direction of the groups behaviour. The researcher should not disturb the setting as it should be as natural as possible. Blending in is very important even though it can be difficult sometimes. Sometimes listening and looking might not be enough and the researcher will have to be more active and gain information by asking questions. Informal unplanned and unstructured ways are used to obtain data and there might be a great deal of hanging around. Recording information can be problematic, many researchers write up information in the evening once they have thought about the days events. The toilet is sometimes a place where many researchers go and quickly write ups their notes once something significant has happened - this relies on memory which can be selective. Participant observation can be a long process that can take up to a year of research or more. It requires dedication, stamina and possibly a double life whereby the person is cut off from friends and family
Covert participant observations?
This is when the sociologist aims to observe a person/group and the sociologist DOES NOT reveal their true identity but pretends to be an authentic member of the group. The sociologist will not tell the participants that they are really a researcher who wishes to observe their behaviour. The researchers identity remains COVERT/COVERED
Eg James Patrick and Glasgow Gangs
Advantages of covert participant observation?
- uncovers what people actually do (true reflection) so the data gained is very valid as it’s more difficult for participants to lie or mislead the sociologist here
- less demand characteristics would be demonstrated by the group of people being observed. The group will not know they are being observed and therefore will be less inclined to change their behaviour and will behave more naturally (high in validity)
- can uncover a great deal of insight and fresh in depth information so verstehen is very high
- a method to successfully study some topics that otherwise could’ve been impossible. Eg criminal gangs, closed off to outsiders and would reject some people, if the researcher can gain their trust rapport and respect then the group might allow access
- can be conducted by a single researcher without them doing a great deal of preparation. This makes the costs of the research quite low.
Disadvantages of covert participant observation?
Can be unethical eg lack of fully informed consent and deception
Could be difficult to take accurate notes without being seen as writing notes down could arouse suspicion and put the researcher at danger especially with a deviant group.
Moral dilemmas eg observing a gang undercover, they might want the researcher to take part in illegal activities and it would require a great deal of effort / acting ability to fit in.
Pose danger to the sociologist eg Ken Pryce was murdered whilst trying to carry out a participant observation of organised drug crime.
Overt participant observation?
This is when the sociologist aims to observe a person/group and the sociologist DOES REVEAL their true identity. The sociologist will tell the participants that they are really a researcher who wishes to observe their behaviour. The sociologists true identity is OPEN and NOT COVERED to some or all of the group who are aware that the researcher is actively observing them.
Advantages of overt participant observation?
Very ethical because participants would’ve given fully informed consent and deception would be minimal
The researcher can freely take notes that will be more accurate and less rushed
The researcher can freely ask lots of questions without participants becoming suspicious (gain verstehen)
Disadvantages of overt participant observation?
Less valid than covert observation
Social desirability bias and demand characteristics might occur which could make the results inaccurate and biased
Lacks verstehen and true insight in to human behaviour
Time consuming as it can take months maybe years
Some observers get attached to the group so there e observations become biased. The observer becomes too sympathetic towards the group and ‘goes native’ - the observer loses detachment/objectivity and identifies too closely with the group.
Why do Positivists criticise participant observation overall?
They question the reliability of the results because the success is generally due to the relationship established between the researcher and the group and is therefore hard to replicate.
Lack of representativeness, because the number of people observed is generally quite small so it’s not possible to generalise from the findings to members of similar groups.
They also say it is unsystematic, unstructured and unreliable because they focus on naturally occurring behaviour.
Non participant or direct observation?
Involves the researcher sitting and observing an activity. They’re detached, objective and an unobtrusive onlooker who plays no active role in the activity being observed.
Is structured using a coded observation schedule that directs what is to be observed eg particular types of behaviour or activity. This is favoured by Positivists as they’re gaining ‘facts’ in the form of quantifiable data.
Non participant observation advantages?
As the researcher is objective and detached, there’s less bias when interpreting the groups behaviour.
And as the researcher is not joining in activities, the group itself shouldn’t be influenced by the observer.
Non participant observation disadvantages?
However,
The observer is likely to be observing artificial behaviour caused by their actual presence eg an oftsted inspector in a classroom will generally affect how both the teacher and students act compared to how they do generally.
It also gives little insight as to why people behave the way that they do, lacks verstehen.
Objectivity may be difficult to achieve as observers have to make value judgements about whether behaviour or activities fit particular categories on their observation scale.
Mixed methods ?
Most sociologist research will mix and match methods in order to guarantee both reliability and validity; very few will focus entirely on the use of either quantitative (positivist) or qualitative (interpretivist) methods alone. There’s great flexibility, as long as the method fits the purpose of research.
There are two ways of doing this, triangulation and methodological pluralism.
Methodological pluralism?
It treats all methods as equal and assesses the strengths and weaknesses of each method in terms of his if fits into the research task. It’s the Combining different research methods in order to build up a fuller picture of what is being studied as it produces a more comprehensive view of social reality and can gain new insights; primary and secondary and quantitative and qualitative.
Triangulation by Denzin 1970?
Involves the use of more than one method or source of data in the study allowing findings to be cross checked. Two or more methods or data collection are used and their findings should correspond with each other and agree. Triangulation uses different methods and
theories to check findings against each other. The quality of the research should be enhanced by using both qualitative and quantifiable data and can therefore enhance reliability and validity.
Triangulation can take different forms?
Data triangulation -
Collecting data at different times from different people in different places as a cross check for validity.
Investigator triangulation -
The use of different researchers, observers and interviewers to check for interviewer bias
Within method triangulation -
Variety of techniques within the same method eg different forms of a questionnaire in order to check the validity of the answers and the reliability of the method itself
Between method triangulation -
A combination of a number of research methods eg questionnaires and interviews etc. To see if the data produced by one method can be checked by comparing it with the data produced by other methods.
Case study - pluralistic approach ?
Detailed and in depth examination of one particular case or instance or something using methodological pluralism. They’re useful because they look at a single thing (eg life history, organisation history, one specific event etc) from several angles using several methods. This means that the sociologist can build up a rich picture of the dynamics that underpin the actions of a particular person, situation or institution.
Building up a rapport
Interpretivists would support the use of
Respondents feel more comfortable speaking the truth and opening up so give better answers which ensures validity
Sensitive topics can also be discussed eg sexual history etc.
Sociologist builds up a level of trust which isn’t possible in questionnaires for example. So u can gain verstehen
Mixed methods evaluation?
+
Great flexibility, not restricted
Good to improve validity
Can improve reliability as more than 1 method can check if results are similar
-
Time consuming as opposed to one method
Can be difficult when each method produces conflicting results
Researcher needs to be very skilled to blend the data in order to answer the one question
