Foundations of Social Science Research Notes

What is the difference between validity and reliability?

Validity - are our results valid for what we are trying to use them for, based on the methodology we used.

Reliability - are results consistently saying the same thing

What is pseudoscience?

When people come up with statements that they say are true and backed by science, but the science is based on non scientific principles and therefore not accurate

What are the 8 common types of thinking errors when it comes to analysing things, both day to day and in research?

1. Overgeneralisation - taking some evidence and over-extrapolating the results

2. Selective observation

3. Premature Closure

4. False Consensus

5. Halo Effect

6. Confirmation Bias

7. Belief perseverance

8. Availability heuristic

(Think Old Sailors Prefer Fresh Hot Coffee Before Anything)

What is overgeneralisation, give an example?

Taking some evidence for something and over-extrapolating the results to something else.

E.g. We see that some homeless people drink alcohol so think all homeless people must be alcoholics

What is selective observation, give an example?

Like confirmation bias, but instead of altering or rejecting things we only look at things which support our theory and ignore all else.

E.g. We think all leafs are green, so we ignore brown and maple leafs.

What is premature closure, give an example?

When we don’t quite finish researching something and decide to draw a conclusion anyway

E.g. we are researching the best microwave, and we’ve looked at a few and find one with excellent reviews and that ticks all our boxes. We don’t keep researching and look at any others and conclude this is the best one

What is false consensus, give an example?

This is where we extrapolate our thinking onto others

E.g. I think that tractors are great machines so everyone else must too adn if they don’t there is something wrong with them

What is halo effect, give an example?

This is where we have a positive association with something, and because of that, we extrapolate the positivity to other things that one thing did

e.g. I think the UN writes excellent special reports into situations, I therefore extrapolate that to every UN report and assume they will all be as high quality, unbiased, and objective. I might also extrapolate this to decisions the UN makes, thinking they will put as much work into these as their reports.

What are the 3 main epistemological frameworks?

1. Positivism

2. Interpretivism

3. Critical

What is belief perseverance? Provide an example.

This is where even though we have evidence against a belief we hold, we maintain this belief anyway. 

E.g. We believe the world is flat, years later overwhelming evidence supports it is round, we however, continue to believe it is flat because we hold the belief so strongly.

What is belief confirmation bias? Provide an example.

This is similar to selective observation, but instead of focusing only on what proves our theory like in selective observation, we look at everything and interpret evidence or change things to fit our theory

E.g. We think that cars are causing climate change, we then start to pay special attention to the evidence that supports this, and change or disregard other evidence. E.g. looking at articles on climate change being caused by vehicle emissions, ignoring any critique’s of this as deniers, and thinking that the article you did read rejecting it was actually more evidence for it, as it didn’t categorically say the cars weren’t causing the climate change.

What is availability heuristic? Provide an example.

This is where we overestimate the amount something does happen due to experiencing a memorable event of it. 

E.g. we see a memorable movie about nuclear disasters, and think they must occur a lot

What makes the research process strong evidence, and a strong way of knowing that something is true?

• Consistently, we use the same term to describe the same thing across literature

• We follow carefully made step by step procedures to gather data and analyse it

• Multiple observers are used to increase objectivity in analysis of what is happening

• Conclusions are all based on data

• Our work is carefully documented and can be repeated with the same results

What is the difference between qualitative and quantitative data?

Both are equally important and useful forms of data, however, are best used in different situations

Quantitative:

• Excellent for when we have little knowledge about why something may be happening, and need to develop an understanding from very little

• Uses words, experiences, anything that isn’t numerically recorded essentialy

• Usually inductive, meaning we come up with new theories from the research rather than apply current ones

• E.g. can be used to find why people think a certain way, what factors usually make up how they feel about something (such as their happiness), what they were thinking, how they understand things, etc

Qualitative:

• Excellent for when we have an idea about how something may be happening, and we want to drill further into what that is.

• Quite specific, and useful for testing specific hypotheses.

• Uses numeric data which is statistically analysed to find meaning

• Usually deductive rather than inductive, meaning we apply the results ot current theory rather than make new theory

• E.g. can be used to test whether homelessness is caused by a certain theory, to measure the effectiveness of interventions against homelessness, to identify different categories of homeless trajectories, etc

What is scientific literacy?

The ability to understand and use scientific processes, principles, and techniques

What is the difference between methodology and methods?

Methodology refers to the research as a whole, what was our idea, what are the ethical considerations, what is our strategy to do this? E.g. are we using qual or quant research designs, what are our research questions, etc

Methods refers to the specific tools and techniques that we use in our investigations to answer the research question and align with our methodology E.g. where are participants coming from, what will we use for data analysis, what type of design are we using

What is an RCT?

Randomised Control Trial/Test

This is where you randomly allocate participants to categories, with at least one control and experimental category to determine the effectiveness of treatments

What does a P value show

What percentage we are pretty sure the data is different from the other category without being by chance

What determines which approach we take to research?

What we are trying to find out, e.g. are we trying to find the nitty gritty answer to a specific question and are linking to existent theory, or are we trying to find information on something we don’t know a lot about, and therefore actually need to make theory on.

What does ontology mean?

Ontology refers to what exists, and what the fundamental categories of reality are

What are the 2 positions to ontology?

Realist and normative

How do normative perspectives see ontology?

A normalist approach to ontology sees what the world is as being dependant on peoples subjective views, and therefore different for different people

(think normalist = normal, and it is normal for people to interpret things differently)

How does a realist perspective see ontology?

A realist sees that what we see in the world is what there is, as these things are objective.

(Think realist = what is real and the approach is all about what is real being what is there)

What is epistomology?

Epistomology looks at how we know what we think exists actually exists. How do we know what we think is happening is happening.

What is a paradigm

A ridiculous word

In reality it is another word for framework, and refers to the framework in which people are looking at things.

What is a positivist framework?

This is a framework which is trying to find the objective truth about one specific thing

Who were key influencers of positivist frameworks and what did they do?

Comte came up with it, and believed the best way to identify if something was true was to continue testing, and if the results consistently proved it, it was likely to be true.

Popper however, refined this theory by saying that a faster method would be to try to disprove something, and if it couldn’t be disproved then it was likely to be true.

What is an interpretivist framework?

This is a framework looking to understand how people feel about things and make sense of their world.

Who was influential in creating interpretivism?

Max Weber came up with interpretivisim because as a social scientist he wanted to understand the nature of our world and what people thought of it. He decided that the meaning and nature of the world was based on peoples subjective experiences. 

Therefore, he said that theories are only true if the people they apply to perceive them to be, and that theories had to be constantly built on to find the “truth”

What is a critical framework?

These are frameworks that combine positvism and interpretivism to expose injustices in marjinalised and oppressed groups and challenge peoples thinking about these groups

What is transferrence?

This is when we transfer our perspective of one thing onto something else.

E.g. we meet someone who reminds us of our Grandfather and consequently transfer his personality onto them by thinking about them in the same way we thought about our Grandfather, and treat them similarly.

What is the difference between analytico inductive and hypothetico deductive reasoning?

Analytico inductive reasoning is where theories are produced based on research, and research has the aim of producing these theories. 

Hypothetical deductive reasoning is where research is applied to existing theories to try to drill down on them and the reasoning behind something.

(think analytico inductive is analysing the situation and coming up with new theory, whereas hypothetico deductive is making a hypothesis about how theory applies to something and deducing whether this is true)

What is the definition of ethics?

Moral standards that govern our behaviour in a certain setting or as part of a certain group.

What is the history of the codes of ethics in research?

The code of ethics started with the Nuremberg trials which sparked the Nuremberg code - stating all research should be for the benefit of humanity (forgot about participants) as a preliminary code of ethics. This was soon followed by the World Medical Assoc Intl Statement, and then the Helsinki Declaration, which became the basis for numerous codes of ethics, including the National Health and Medical Research Council’s Code of Ethics.

What is frivolous research?

Research that is performed for no benefit/purpose, this is unethical since all research has risk and creates negative impacts

What are the core four areas of the NHMRC code of conduct for human research?

1. Justice

2. Respect for human beings

3. Beneficence

4. Merit and Integrity

What is in the core area of Justice regarding the NHMRC research code of conduct?

Distributive justice:
Research benefits people and is distributed to be of benefit to toehrs

Procedural Justice:
Participants are treated fairly, and equitably when required

What is in the core area of Beneficence regarding the NHMRC research code of conduct?

The benefits of research outweighs the risks of performing it.

What is in the core area of merit and integrity regarding the NHMRC research code of conduct?

Merit:
- Research that is beneficial, which uses a sound methodology, which is based on previous literature, is supervised appropriately, respects participants, and uses appropriate facilities

Integrity:

This means researchers are researching with the genuine motivation of unlocking new knowledge, while being honest, following codes of conduct, and presenting results effectively and scrutinisably. 

It also means research that is not frivoulous.

What is a meta analysis?

This is an analysis of the results lots of literature in an area, which is evaluated in relation to a research question.

What is a battery survey?

This is a big survey with smaller survey inside of it.

E.g. a survey that includes a depression survey and an intrusive thoughts survey

What is the difference between ethics and morals?

Ethics are standardised morals we agree to
Morals are just the individual ways we feel we should act

What are the 5 core principles of ethical conduct which stem from the NHMRC’s code of ethics?

1. Do no harm - includes all types of harm including legal, financial, mental, physical, etc

2. Voluntary Participation - Ensuring participants are free to join, decline, or leave without adverse effects

3. Informed Consent - Participants are informed about the study and know they are free to leave and consent to partake in it

4. Avoiding deceit - Ensuring participants are not deceived, and that if they are, it is necessary, they are safe, and they are fully debriefed afterward

5. Confidentiality/Anonymity - Ensuring participants are not identifiable by researchers or if there is no way for them not to be, that what they say will not be traceable back to them for others who read it

Think Dangerous Vehicles In A Collission - DVIAC

Within Epistemology, what are the 3 core frameworks of thinking? What do these mean?

1. Positivist approach - usually quant, hypothetico deductive, verifiable objective facts

2. Inerpretivist - usually qual, analytico inductive, looking to understand the meaning people apply to things

3. Critical - mix of both, looking to identify structures causing oppression, expose injustice, and give a voice to oppressed groups

What is Verstehen, and how does it apply to social science research?

Verstehen is understanding how people view their everyday experiences, and is the foundation of interpretivism

What are the 3 ways to conduct quantitative research?

1. Lab studies

2. Field Research

3. Natural Experiments

What are lab studies?

Where research is conducted in regulated conditions designed to simulate the natural environment while controlling for confounding and extraneous variables as much as possible

What is field research?

Where experiments are conducted in the natural setting, but extraneous and confoudning variables are controlled as much as possible

What are natural experiments?

When variables are not manipulated and something is studied naturally occurring (also called a quasi or non-experimental design)

What is randomised sampling?

Where all people in a population have an equal chance of being selected

What is systematic sampling?

Same as random, but where every xth person is chosen, x can be any so number e.g. every 2nd or 6th person is chosen from the population

What is convenience sampling?

Where a group of people who meet the requirements for the study are taken from a specific spot(s)

What is stratified sampling?

Where you look at a population, and make sure the proportions of certain subgroups represent the population in your sample

What is quota sampling?

The same as stratified sampling, but where this is repeated until a certain number of responses have been recorded

What is snowball sampling?

Where participants have to send another person to do the study once they have finished

What is cluster sampling?

Where a random selection of people with something in common is chosen

What can internal validity be influenced by?

1. Confounding variables

2. Whether there was a control group and whether this did its job properly

What can external validity be influenced by?

1. Whether participants knew they were being observed

2. Whether results are generalisable

What demonstrates a study has strong reliability

Results are consistent over time

What is nominal data?

Categories that can’t be put into an order

What is ordninal data?

Categories that can be put into an order, but the distances between them are unknown

What is interval data?

Categories that can be put into an order but the gaps between them are all the same size

What is ratio data?

Categories that can be put into an order, and 0 means 0

What is discrete vs continuous data?

Discrete data is whole numbers, whereas continuous is data with decimals

What is a frequency distribution?

How many times different aspects of a variable occur:

e.g. if looking at a scale, how many people ranked something as 5/10, 6/10, 1/10, 3/10, etc for the whole scale

What is bootstrapping?

Filling in blanks in responses

What does Standard Deviation (SD) show

SD shows how much the data deviates from the average

What is the 68, 95, 99.7 rule and what does it mean?

This is a rule for analysing SD values, it means that:
1. 68% of the data will land within 1 SD of the mean
2. 95% of the data will land within 2 SD’s of the mean
3. 99.7% of the data will land within 3 SD’s of the mean

What is a subgroup?

A group within a group e.g. a group would be ethnicity and subgroups would be the type of ethnicity, this could be more specific or less specific depending on the starting group

What is a likert scale and what on earth are likerts?

A likert scale is one of those scales with equal points on them, e.g. from one to ten, etc. For a likert scale to be a likert scale it has to have an odd number of points, as one point has to be a neutral option in the middle

A likert is just a point on a likert scale

What does it mean to collapse data?

This means to break the data into different parts, which is done to represent different things

e.g. breaking it into male and female to compare scores statistically

When was the Nuremburg code created?

1947

When was the Helsinki Declaration first adopted?

1964

What are 4 different methods for qualitative studies?

1. In depth interviews

2. Ethnographic Studies

3. Focus Groups

4. Participant Observation

What are in depth interviews?

These go beyond surface level on a certain topic, and can be structured, semi-structured, or unstructured

What are ethnographic studies?

These are where a researcher places themselves within the environment of the population they are interested in to understand the dynamics

e.g. this could be placing yourself in a gang to understand gang culture

What are focus groups?

This is where a group of people with certain traits come together to discuss a certain topic with you

What is participant observation?

Observing what people do in a certain situation

What is the difference between structured, semistructured, and unstructured interviews?

Structured - certain questions and you can’t deviate from these

Semi-Structured - certain questions with room for you to add as necessary to get additional information

Unstructured - Might have a starting point and then go where-ever the interview takes you

What is a narrative review?

Looking at building the story for how people view or see certain things

e.g. what do homeless people think causes homelessness

What is transcribing and where is it used

Transcribing is where the information from a qualitative interview is recorded verbatim (word for word exactly as it was said)

What is the difference between Jeffersionian and Orthographic transcription?

Orthographic - Transcribing word for word exactly

Jeffersionian - Transcribing based on themes

Since qualitative research doesn’t rely so much on experimental settings to measure things, what are considerations to make to ensure reliable and valid research?

The main consideration with qualitative research is how we conduct the interviews, e.g. we want to:

• Explore inconsistencies but not force consistency

• Ensure the data collected during each interview is comprehensive 

• Allow participants to lead and talk the most

• Be well organised to avoid confusing participants when going back and forth

• Ensure questions are well designed and phrased

• Ask for elaboration as appropriate

What is thematic analysis, and how do you do it?

This is one way of analysing qualitative research, where you:

1. Familiarise yourself with the data

2. Come up with codes (like different colour highlighting) for different themes in the data

3. Sort the data based on the themes you identified, and identify an overarching theme

4. Review the themes, defining them and naming them

5. Write the report

All Done!