quantitative and qualitative methods

Question 1

overview

Answer 1

• split approaches to research into 2 broad categories
• quantitative methods - collect numbers/numerical data and use statistical tools
• qualitative methods - collect words, pictures and artefacts
• some researchers also adopt both approaches or apply quantitative methods to qualitative style data
• quantitative methods are probably easier to group together, because qualitative methods include such a broad range of approaches.

Question 2

outline of quantitative methods

Answer 2

• take a phenomenon and try to condense it down into a few dimensions or variables that can be measured as precisely and reliable as possible.
• very important to choose variables that are representative of the phenomenon you’re studying.
• choosing variables that are representative of the phenomenon you’re interested in involves operationalisation.
• operationalism means choosing a measurable proxy for the phenomenon you’re interested in
• often make use of statistical methods
• using statistical methods means looking at lots of cases
• the goal is often to develop generalisations, or theories that are generally applicable.
• involves testing predictions that logically follow from theories.

Question 3

outline qualitative methods

Answer 3

• are focused on meaning rather than measurement
• instead of condensing a phenomenon down to a simple set of features or dimensions, qualitative research tries to examine many features.
• try to look at all aspects of one or few instances of a phenomenon
• view the context as a central part of the phenomenon being studied
• also emphasise the idea of following data wherever it leads

Question 4

qualitative methods

Answer 4

• extremely varied with different methodologies, underlying theoretical assumptions, and intellectual histories.
• examples - verbal protocol analysis, ethnographic methods, discourse analysis, phenomenology
• there are many more, including case studies, grounded theory, participatory research, focus groups.

Question 5

verbal protocol analysis

Answer 5

• also known as thinking-aloud protocols
• involves collecting and analysing verbal data on cognitive processing
• pps are given a task and are asked to verbalise what they are thinking as they go about solving the task
• data are coded and analysed to infer the information processing steps involved in solving the problem.
• the approach was used in early cognitive science by Simon and Newell who were pioneering researchers in cognitive science and artificial intelligence
• carries certain assumptions about the nature of human cognition/thinking

Question 6

ethnography

Answer 6

• more a style of research than a method of data collection, it involves studying people in “the field”, and requires the researcher to enter the setting they are studying.
• attempts to understand how the socio-cultural practices and behaviours of people are shaped by their social, physical and cultural contexts.
• tries to make sense of events from the perspective of the pps
• could include data from interviews, or participant observation
• in cognitive psych, ethnographic approaches have been used to understand how people style problems in real-world settings.
• in critical psych, ethnographic approaches have been used to understand the interplay between race, class, gender and education in shaping participants life worlds.

Question 7

discourse analysis

Answer 7

• the social study of language used in talk, text and other forms of communication
• it involves a distinctive way of thinking about talk and text where language doesn’t just represent the world but also contracts the world.
• strengths - allows you to examine how language constructs reality, and can make use of primary data, or secondary data
• can be difficult to use discourse analysis to develop the same kind of generalisations as you might develop with other approaches.

Question 8

phenomenology

Answer 8

• associated with philosophers Husserl, Merleau-Ponty and Sartre
• involves bracketing off any preconceived notions we might have about a phenomenon to achieve an understanding of that phenomenon that has not been influences by our prior beliefs.
• emphasises people’s first-hand experience and attempts to understand and describe subjective experience from the pps point of view.
• used in fields like cognitive psych to understand the nature of subjective sensory experiences.

Question 9

issues in qualitative research

Answer 9

• unlike quantitative methods that might use printed questionnaires or computers to record any measure response, in qualitative r4esearch, the researcher is the instrument
• important for researchers to reflect on their values, assumptions, biases and beliefs to understand how these might impact the research
• the research instrument can change, e.g., in ethnographic research the changes in the researchers experience might alter how they record and observe behaviours.
• there are parallels to validity, reliability and objectivity in qualitative research
• these are:
  1) credibility - Can the data support the claims. Can be established through prolonged engagement, discussions with other researchers/participants, and critical self-reflection
  2) transferability - Can the findings be transferred to similar contexts. Requires extensive, detailed, and careful descriptions of the research context (“thick descriptions”).
  3) dependability - Ensuring that researchers maintain a record of changes in the research process or research instrument (i.e., themselves) over time.
  4) confirmability - Concerned with ensuring that the data used to support the conclusions are verifiable.

Question 10

quantitative methods

Answer 10

• putting numbers to thing so that it can be measured
• the motivation behind measuring phenomena is that measurements are publicly available and verifiable
• unlike qualitative research where researchers try ti simultaneously study many aspects of a single phenomenon, quantitative research tries to condense a phenomenon dow into a single dimensions.
• the first step in quantitative research is often figuring out how to quantify the phenomenon of interest, this involves choosing a proxy that can stand-in for the phenomenon.

Question 11

operationalism

Answer 11

• the process of choosing a proxy is called operationalisation
• ways to choose a measurable proxy is - develop a scale or questionnaire
• measurements have to be reliable and valid

Question 12

quantitative methods and causation

Answer 12

• qualitative research studies phenomenon in the wild, but quantitative approaches instead try to exert a lot of control over phenomena
• control allows researchers to make claims about causation ad give casual explanations
• few ways to understand causation:
  1) as a difference that makes a difference
  2) in terms of manipulation
  3) in terms of probability

Question 13

what is a cause?

Answer 13

• one view of causation can be summed up as a difference that makes a difference
• if you take two situations, one in which the phenomenon occurs and another in which does not occur then whatever is different between those situations is the cause of the phenomenon.
• you can also understand causation in terms of manipulation - if you can manipulate one thing and observe a change in another, then the two things may be casually connected.
• by intervening and manipulating parts of a system you can identify how they work
• causation can also be understood in terms of probability - if the presence of one thing increases the probability of the other thing occurring, then there may be a casual relationship

Question 14

causation and confounds

Answer 14

• to be justified in claiming a casual relationship other conditions must usually be met
• but casual claims are not always black and white, sometimes we can only be more or less sure about casual relationships

Question 15

smoking and confounds

Answer 15

• let’s say you are studying the relationship between emphysema and cancer, because you think emphysema might cause cancer
• in this situation, smoking is a confound
• if you wanted to see whether emphysema caused cancer then you’d have to control for smoking
• only look at smokers and see if there’s still a relationship between emphysema and cancer or whether cancer also occurs in the absence of emphysema
• only look at non-smokers and see whether emphysema and cancer are still related or whether cancer develops in the absence of emphysema
• emphysema and cancer are correlated (the increase in one leads to an increase in the other), but emphysema doesn’t cause cancer because they have a common cause.

Question 16

correlation and causation

Answer 16

• sometimes two correlated variables have a casual relationship - smoking and cancer
• sometimes they have a common cause - emphysema and cancer
• sometimes they have neither - from 2000 to 2009, there was a string relationship between the number of men getting engineering degrees and per capita consumption of sour cream.

Question 17

qualitative vs quantitative methods

Answer 17

• in qualitative research, you study phenomena in context while in quantitative research you aim for control.
• but you can use either approach to study the same phenomena/psychological processes

Question 18

quantitative study of memory

Answer 18

• you could use experiments in a lab where you give people lists of words to remember
• you could manipulate aspects of the worlds, and measure performance to try to understand something about memory and emotional salience
• ensure that the only thing that differs between the words on each list is the emotional salience
• control for possible confounds like:
  1) word length - make sure that one list doesn’t contain ling words and the other short words
  2) word order - make sure some people get the lists in one order and some in the other order, because maybe people get tired by the end and that influences memory

Question 19

qualitative study of memory

Answer 19

• for a qualitative approach you don’t want to study memory in the lab - you want to study it in the wild - this allows you to ask different kinds of questions.
• you could use an ethnographic approach with, for example, bartenders:
  1) you might do fieldwork in a bar observing bartenders
  2) through this, you might see that bartenders structure their environment in a particular way
• this might lead you to form the hypothesis that bartenders structure their environment to support their memory
  3) conduct follow-up interviews or observe the training of bartenders for further evidence for
  4) engage in bartending and critically reflect on your own experience

Question 20

computer simulation and formal methods

Answer 20

• qualitative and quantitative methods try to understand phenomena by studying the phenomena themselves - the data they use comes from the phenomena
• in approaches like computed simulation and formal/mathematical modelling researchers instead generate the data
• researchers try to build systems that replicate or reproduce some aspects of systems or phenomena they are studying.
• this might allow them to gain new insights into these systems
• comparing the behaviour of their artificial systems with the neutral system allows researchers to test theories about the processes that produce phenomena.

Question 21

computer simulation

Answer 21

• It has been used to study a lot of different phenomena in psychology.
• Simulation has been used to show how seemingly complex behaviour can arise from very simple processes.
• Flocking behaviours in birds seems very complex, and it looks as if there must be something very complex going on inside their brains.
• But you can simulate this behaviour with only three simple rules:
  1) Avoid collisions with other birds.
  2) Align direction with nearby birds.
  3) Approach distinct birds.

Question 22

agent-based modelling (ABM)

Answer 22

• Agent-based modelling tasks a cue from approaches like those used to model bird flocking and Conway’s Game of Life.
• In an agent-based model, the researcher stimulates a group of ‘agents’.
  1) The ‘agents’ will typically have some memory, a set of goals, and some rules.
  2) The memory allows them to store their current state or consequences of their previous actions.
  3) The goals usually represent some state they’re trying to achieve.
  4) And rules govern their interactions.
• By allowing these agents to interact, and by manipulating aspects of the agents, it’s possible to see how social phenomena can arise.
• Agent-based modelling can be used for modelling phenomena like the spread of misinformation through social groups:
  5) If you thought misinformation was more likely to spread if passed on by particularly influential individuals, then you could include this idea in your simulation.
  6) Or if you thought that misinformation was more likely to spread inside socially isolated groups, then you could modify your simulation to create socially isolated groups to test this hypothesis.
• After this simulation, you could still go and check the real world to see if it behaves like your simulation.