Descriptive Methods, Qualitative Methods & Sampling Flashcards
Hey Sam
Just here to remind you that you are incredibly amazing, I love you heaps and you make me very happy. You’re a truly good person and it’s impossible not to smile when you’re around
Non-experimental design
These methods allow you to describe a situation but don’t allow predictions or establish cause-and-effect.
Observational method
The basic form is observing human or animal behaviour. Not used often in psychology because we usually want to do more than ‘describe’ but it is a useful starting point. Also, some behaviours can’t be studied experimentally such as illegal behaviours.
The two types of observational studies
- Naturalistic / Field
2. Laboratory / Systematic
Naturalistic Observation
Good for observing “natural” or true behaviours. Such as eavesdropping. Observe people/animals in their natural habitat which increases ecological validity (the extent to which research can be generalised to real-life situations.
The options of Naturalistic Observation
- Undisguised observation: Participants aware they’re being watched
- Disguised observation: Ps unaware they’re being watched
- Participant observation: Researcher does participate in situation
- Nonparticipant observation: Researcher doesn’t participate in situation in which Ps are involved
Laboratory Observation
Good for observing behaviour in a “contrived” setting such as the Strange Situation Test. It focuses on fewer behaviours, the participants often know they will be observed but there are exceptions such as the Strange Situation. Experimenters can be participant or nonparticipant, disguised or undisguised. For example the Strange Situation Test is undisguised for mothers, but disguised for children, nonparticipant observation.
Potential problems of Laboratory observation
Undisguised observations may have a reactivity problem where they may simply react to being in a lab setting.
There are expectancy effects (is the researcher just paying attention to behaviours that support hypotheses).
Advantages of laboratory observation
Reduces time needed to conduct the study
Reduces cost
What are the means of data collection in observational studies
Narrative records
Checklists
Narrative records
Can be hand-written notes or video-tape (& notes taken later). Purpose is to capture everything a participant said or did during specified time period. Can be subjective (e.g., expectancy effects). Data must be coded in order to be analysed (time-consuming). Should be coded by more than one person (inter-rater reliability).
Checklists
A tally sheet where the researcher records P’s attributes or whether behaviours were observed. There are static attributes which do not change and action attributes that change based on the participants behaviour. Researchers must be clear on operational definitions of behaviours they plan to observe
Qualitative Methods
Social research uses observations / unstructured interviews in natural settings. Qualitative data is analysed / interpreted without statistics. Qualitative researchers are interested in the subjective experiences of people. Quantitative researchers begin with an aim / hypothesis and then make observations. Qualitative researchers make observations and, in the course of doing so, may develop hypotheses
Characteristics of Qualitative researchers
Expect unknowns
Flexible skill set
Resourceful
Natural environments
Interested in quality measurement
Characteristics of quantitative researchers
Maintain detachment
Value experimental controls and hypothesis testing
Interested in quantity measurement
“4 R’s criteria” for assessing quality, validity and value of qualitative research:
Rigour
Relevance
Resonance
Reflexivity
Rigour
Has the research been carried out competently?
Are the interpretations suitably justified and plausible?
Relevance
What contribution does the research make to the existing body of knowledge?
Resonance
Do the findings of their search ‘speak’ to the reader?
Does it resonate with many people’s experiences?
Reflexivity
Adopting an attitude of self-awareness & active reflection throughout the project > How do you influence the process?
Case Study Method
Comprises an “in-depth” study of one or a few individuals, groups, social settings or events. Provides foundation for future studies (i.e., hypothesis generating). Provides a method to study rare phenomena (e.g., split-brain patients).
Problems with the case study method
External validity: Individuals observed may be atypical (i.e., unrepresentative of population). Generalisations to population could be erroneous
Expectancy effects: Biased interpretation of observations or data collection
Archival method
Describing data that already exists. E.g. using case reports from welfare agencies to understand family violence. Many sources can be used such as the ABS.
Benefits of archival method
Low cost
No need to recruit/observe participants
Minimises reactivity
Less time consuming (because data already exist)
Problems of archival method
Low control
Hard to assess whether data are reliable/valid
Does archive provide ALL data that was collected?
Interviews
Typically involves asking questions in a face-to-face manner. They can be conducted anywhere.
3 types of interviews
Standardised
Semi-standardised
Unstructured
Standardised interview
Formal structure
Qs asked in specific order
Qs are read as they’re written
No explanation provided
Semi-standardised interview
Some structure but wording of Qs is flexible.
Language can be modified to accommodate person.
Interviewer can provide further explanation.
Interviewer can add/delete Qs.
Unstructured interview
No set order and no specific wording. Qs are spontaneous and free-flowing
Pros of interviews
Can record non-verbal communication
Encourage responding
Cons of interviews
Time consuming
Expensive
Interviewer bias (tendency for person asking the questions to bias the participant’s responses)
Action research
Research conducted by a group to identify a problem (phase 1), attempt to resolve it (phase 2) and assess whether the resolution was successful (phase 3).
Surveys
A set of self-report items that is completed by respondents at their own pace, often without supervision. Questions should be clear, precise and use simple language. They can be open ended, closed-ended, partially open or rating scale questions.
Pros of surveys
Honesty due to perceived anonymity
Interviewing skill not relevant
Cons of surveys
Questions can be emotionally laden or loaded. Leading questions can influence answers. Double-barrelled questions can be misleading (asking multiple things in one question). There cann be a response bias (such as agreeing o disagreeing with everything)
Methods of undertaking surveys
Email surveys cam eliminate interviewer bias however it can lead to a low response rate.
Telephone surveys are more time consuming and costly but questions can be clarified.
Personal interviews can also be conducted however these have quite a lot of interviewer bias.
Descriptive methods
Observational methods (naturalistic and laboratory)
Qualitative methods list
Case studies
Archival methods
Interviews
Action research
Surveys
Population
The entire group of people about whom a researcher wants to learn.
Census
A survey of the entire population. Sampling the population is often impractical/impossible.
Sampling
Sampling the population is often impractical/impossible. We select a sample of individuals (or objects) from the target population and inferences are made about the population based on this sample
Sample
The group of people who actually participate in a research project
Statistical Inference
The process by which we derive generalisations about populations on the basis of sample data. Our goal is to generalise findings from a sample to the population. For these conclusions to be reliable and valid, it is important that our sample closely represents the population.
Representative sample
A sample that is approximately the same as the population in every respect.
Probability Sampling
Sampling procedures used to ensure that each person in a population has a known chance of being selected to be part of the sample.
Non-probability Sampling
Methods in which the odds of selecting a particular individual are not known.
Types of probability sampling
Simple random sampling
Systematic random sampling
Stratified sampling
Cluster Sampling
Simple random sampling
Every person in the population has an equal chance of being included in the sample. A sampling frame can be used (a list indicating an entire population). Random selection can also be used such as random number tables, computer programs and the internet.
Systematic random sampling
Involves selecting every nth person from a sampling frame
Stratified sampling
Involves dividing a sample into subgroups (or strata) and then selecting samples from each of these groups. There are two methods, proportionate and disproportionate.
Strata
Population subgroups used in stratified sampling. Such as men and women in Fed Uni Psych
Proportionate stratified sampling
Is stratified sampling where the sample size proportions for each subgroup correspond to each subgroup’s proportion in the population.
Disproportionate stratified sampling
Is stratified sampling where the sample size proportions for each subgroup do not correspond to each subgroup’s proportion in the population
Oversampling
A procedure used in stratified sampling in which a greater proportion of individuals are sampled from some strata than from others.
Cluster sampling
A probability sampling technique in which a population is divided into groups (clusters) for which there are sampling frames and then some of the clusters are chosen to be sampled.
Sampling bias
Occurs when a sample is not actually representative of the population because the probability with which members of the population have been selected for participation is not known.
If the characteristics of the sample do not approximate that of the population, it is said to be a biased sample e.g., if I only have Caucasian Australian males i`n my sample then it is a biased sample.
Non-probability sampling methods drawbacks
Not every individual in the population has an equal probability of being selected. Thus, it is not a random process, and is less likely to result in a representative sample
Non-probability sampling methods advantages
It is not always possible to obtain all the information necessary about a population to employ a probability sampling method. Non-probability sampling methods are much easier to use.
Non-probability sampling methods list
Convenience samples
Quota sampling
Snowball sampling
Convenience samples
Non-probability samples containing whatever individuals are readily available, with no attempt to make the sample representative of the population e.g.,1st year Psychology participant pool.
Quota sampling
Identifies subgroups within the population on certain characteristics, but uses convenience sampling to obtain the participants. i.e., quotas are set for each subgroup. Once quota for subgroup is met, no other individuals from that subgroup will be recruited. For example 45 females from various age groups (e.g., under 25, over 65, and so on),
Snowball sampling
A non-probabilistic sampling technique in which one or more members of a population are located and used to lead the researchers to other members of the population. e.g., participants might pass on slips of paper with details. Good when looking for an obscure population.
Sample Size
Small sample sizes = big difference between population’s mean & sample’s mean. e.g., when sample size is 2, average distance between sample mean and pop mean is 5. If you increase sample size, there is likely to be less difference between pop’s mean and sample’s mean. Therefore, the larger the sample, the more accurately it represents the population
The rule of thumb in regards to sample size
There are practical limitations that can make it difficult to achieve a large sample size so a compromise is needed. Rule of thumb; aim to have a minimum of 25-30 individuals for each group or condition. This depends on other things too like how big treatment effects are, and exactly what sort of statistics you will be using.
