Research Methods Flashcards
describe naturalistic observation
observing a situation without changing the situation at all. An example of this is David Attenborough observations of animals. This approach has high ecological validity.
describe controlled observation
The observer/researcher change/controls the situation he is observing. An example of this is Bandura’s bobo doll study, researching social learning theory. However, experiments using this approach can lack internal validity (as in demand characteristics or population validity).
describe covert observation
When the participants are unaware that they are being observed. Often reduces demand characteristics. Can cause ethical issues as it could be classed as deception or lack consent. However, according to the BPS code of ethics, they would have to be debriefed of the study, assured of confidentiality and if they were unhappy or upset by the observation, extra procedures would have to take place.
describe overt observation
Participants are aware they are being watched. Can cause the hawthorne or the F.U. effect.
describe participant observation
When the researcher/observer is part of the group they are observing (overt), or if the observer is a confederate pretending to be a part of the group (covert). Examples of this could be if an observer went to live with a special group of people (e.g. a indigenous tribe) temporarily, and recorded their interactions with them. An example of the covert kind of participant observation would be if, for example, a group of individuals pretended to be ill, so they could be admitted to hospital, so they could observe how doctors interact with patients.
describe non-participant observation
If the observer watches from a distance, or observers the actions of other’s interactions with other people. An example of this could be watching people through a one-way mirror or watching how two different people interact as a third party.
What does the mnemonic NOIR mean in terms of data?
N- Nominal Data O - Ordinal Data I - Interval Data R - Ratio Data These data forms go up in accuracy (ratio data is often more accurate or reliable than nominal data)
Describe Nominal Data
Data that relies on named categories e.g. results from a survey asking whether you prefer the sun or the daily telegraph, often presented in bar or pie charts.
Describe Ordinal Data
The levels/conditions of a data can be put in an order, e.g. highest to lowest, but are not necessarily numerical. For example what grades people got at GCSE, it goes from order A-F, but it doesn’t really have a numerical value.
Describe Interval Data
Data that is measured on a scale, ordered categories that are subdivided, so it is not just numbers, but uses numerical intervals. Examples of this include centimetres or litres ( like how many litres of water different people in a sample drink a day).
Describe Ratio Data
Continuous data with no categories. There can be no negative numerical data in ratio data, and though it is similar to interval data, the intervals between readings can be split into ratios, e.g. percentages or degrees celcius.
Describe parametric testing
The data in parametric testing must be either interval or ratio data. The sample data must also show normal distribution and must have homogeneity of variance (e.g. it mustn’t have any random spikes or anomolous results).
Describe non-parametric testing
Used when the sample data does not fill the conditions for parametric tests. Often regarded as less sensitive and powerful to specifics. This is because, unlike parametric tests, they require transformation of data.
What are Heuristics?
Heuristics are how your brain copes with making decisions and avoiding decisions and avoiding decision fatigue by not having to deliberate our choices. It is like a schema we apply instead of deliberating or making a hard decision, e.g. having a staple outfit we repeat instead of choosing one in the morning or ordering something we know we’ve liked before in the at a restaurant rather than having to choose something new from the menu.
Describe Availability heuristics
When the brain naturally determines that the likelihood of something is related to its availability. i.e. believing plane crashes are very likely, because we see them a lot on the news and television, when in fact they are incredibly unlikely. This can be used in marketing.
Describe representative heuristics
The brain’s tendency to focus on the individual rather than the bigger picture i.e. being afraid f being struck by lightning in a thunderstorm, even though there is only a 1 in 700 000 chance, or buying lottery tickets/ being disappointed by losing the lottery, despite there only being about a 1 in 14 million chance of winning.
What is framing?
The way information is presented to determine its effect on whether it is successful or not within consumers, for instance framing it in a positive light so more people are likely to buy it.
Describe how the psychological “foot in the door” technique is used in marketing
The target is given “low stakes” or obvious-answer/emotive questions ( such as “You don’t like burning down rainforests” or “Do you care that most rainforests will be depleted in 12 years?”)
They will then be asked to buy the product, and will be more likely to say yes due to coherence heuristics (the idea that once we agree with something once, it’s difficult to change our mind, even if the information changes).
Describe the key features of a laboratory experiment
- Takes place in highly controlled environments
- Researcher controls the independent variable and records/controls the dependant variable
- The experiments often have low mundane realism and ecological validity
Example = the stanford prison experiment
Describe the key features of a field experiment
- These experiments take place in a neutral, everyday setting
- Researcher manipulates the Independent variable and records the effect on the dependant variable (if there is any, the DV is not as controlled as in lab experiments)
-Often has good mundane realism, population validity and ecological validity as the experiment is usually with the general public or in a public place.
Example = That experiment where different races gender and minorities/stereotypes of people were lying in a town square, pretending to be in pain, and they observed how many (unknowing) participants helped the different confederates.
Describe the key features of a Quasi experiment
- can be considered not a true experiment
- The independent variable is based on an existing difference between the participants (e.g. age/gender), so therefor no one manipulates the independent variable, as having different ages and genders etc. in a sample occurs naturally
- Very common, an example would be the numerous studies completed to try and decipher whether men or women are more intelligent.
Describe the key features of a natural experiment
- The researcher takes advantage of a pre-existing independent variable
- Therefore, the independent variable is not controlled, it occurs naturally
- Often used when manipulating the independent variable is unethical
Describe or draw the table you use to remember whether the IV and/or DV are controlled in an experiment
IV DV
______________________________________
L ✓ ✓
F ✓ X Q X ✓ N X X
describe Pavlov’s experimental research
Pavlov is credited with discovering Classical Conditioning. He did this through experimentation involving dogs and their salivary reflex. He discovered how to turn a neutral stimulus into an cs by encouraging an conditioned response that was originally an unconditioned response to a priorly unconditioned response to and unconditioned stimulus. In this case, the neutral stimulus was the ringing of a bell, the unconditioned response was the dog salivating and the unconditioned stimulus being food. The neutral stimulus originally produced no salivation, while the food (UCS) did produce salivation. Pavlov then rung the bell every time he bought the animal food. The dog then began to salivate at the sound of the bell, meaning it became a conditioned stimulus.
What is an opportunity sample? What are its strengths and weaknesses?
An opportunity sample is a sample using the most convenient or available participant sample, i.e. the nearest group of participants who fit he criteria, or people who walk past you on the street. This is good as it is easy and takes considerably less time than other sampling methods. However, the sample is very susceptible to population bias.
What is a Random Sample? What are its strengths and weaknesses?
A sample gathered either through the lottery method (drawing names out of a hat) or random number table (blindly selecting numbered participants from a table) or a random number generator. The positives of this are that it is unbiased, however it takes time as it needs to have a list of the population being tested and the selection methods can also be time consuming.
What is a stratified Sample? What are its strengths and weaknesses?
When sampling, you split the population into relevant subgroups (strata) (e.g. boys or girls, or age groups) and participants are obtained from each of the different strata in proportion to their occurrence in the population. Selection from the strata is done using a random technique.
What is a Systematic Sample? What are its strengths and weaknesses?
When you use a predetermined system to select participants, such as selecting every nth person from a phonebook (where n = any number). This could be considered unbiased, but also be considered not truly unbiased as it is not truly random unless you use a randomized method to select the starting number/starting participant.
What is a Volunteer Sample? What are its strengths and weaknesses?
Usually advertised in a newspaper or on a noticeboard, often advertised in university campus with small compensation. Can cause volunteer bias.
Name three types of self report techniques
Questionnaires, A structured interview and an unstructured interview.
What is a structured interview?
A structured interview has pre-determined questions, it is essentially a questionnaire that is delivered face to face with no deviation from the original questions.
What is an unstructured interview?
Similar to a structured interview, except new questions are developed during the course of the interview, as well as the pre-determined questions. These additional questions are usually developed based on the answers given by the participant.
What are the three guiding principals when writing questions for questionnaires?
Clarity, Bias and Analysis (CBA 🤦♀️)
What does good clarity look like in a question on a questionnaire?
Questions need to be easy to understand. This means not asking questions that could have two meanings, uses of double negatives and double barrelled questions (questions that ask for two separate answer in one sentence).
What does Bias mean when referring to a good/bad question in a questionnaire?
A biased question might make the respondent more likely to give a particular answer. An example of specific bias that commonly occurs when answering questionnaires would be social desirability bias. This would involve the respondents giving answers in order to look more attractive, generous or generally nice, rather than being truthful. A well designed question would avoid this.
What does analysis mean in the context of creating a good or bad question for a questionnaire?
Questions need to be written so that the answers are easy to analyse. For instance, closed questions can be used to make the questionnaire easier to analyse. These involve only giving a few set answers that the respondent could give to a question, rather than an open question (I.e.what do you like most about your job?), which could give different answers from each participant and therefore give results that are hard to analyse and make inferences from. However, closed questions might give answers that don’t truly represent the respondents real thoughts and
What are four important factors for writing a good questionnaire?
Filler questions, Sequence for the questions, Sampling Technique and Pilot Study.
Why are filler questions important for writing a good questionnaire?
Filler questions can be used to include irrelevant topics to distract the respondent from the main purpose of the survey, therefore reducing the chances of demand characteristics (the Hawthorne and F.U. effect).
Why is the sequence of questions important in making a good survey?
It is best, during the questionnaire, to start with the easiest, saving the questions that are more pressing, so might cause feelings anxiety or a defensive attitude until the respondent has got used to the questioning and feels more comfortable.
Why is sampling technique important in creating a good questionnaire?
The sample used to select respondents, as different sub-groups, such as age and gender, could cause different results.
How would using a pilot study help in the creation if a good questionnaire?
By first testing a draft questionnaire on a smaller sample group, the questions can later be refined in response to any errors or difficulties observed in the pilot study.
What is the ethical issue of informed consent?
4 answers
+ Means the researchers have revealed the true aims of the study and the subject still participates.
+ Has drawbacks, knowing the hypothesis/aims can change the participants behavior
+ Considered a basic human right
+ Also involves researchers revealing any risks or benefits the study might cause (though these can be hard or impossible to predict)
What is the ethical issue of deception?
4 answers
+ Deception can be necessary in research studies when the participant knowing all details of the experiment could change their behavior, therefore affecting the results
+ Considered more reasonable to withhold some of the specifics of the study rather than deliberately giving the participant false information.
+ Can prevent the participant from giving informed consent.
+Could put people off future participation in studies or even psychologists altogether
What is the ethical issue involving the right to withdraw?
3 answers
+ Meaning that Participants know they can, and sometimes do, leave the study at any time
+ Can caused a biased sample for those remaining in the study as they are more likely to be obedient or to conform
+ can be compromised by paid experiments, as participants feel it is less acceptable to withdraw.
What is the ethical issue of protection from harm?
4 answers
+ More important studies tend to have higher risk of distress
+ Often difficult to predict any harm that may be caused, so protection can not be guaranteed.
+ Considered acceptable if the risk of harm is no higher than the risk of ordinary life
+ Also considered acceptable if the individual has been informed and agreed to the risk, or if they leave in roughly the study the same state they came in.
What does the ethical issue of confidentiality entail?
4 answers
+ May be compromised if the researcher wishes to publish the findings.
+ Can be supported by guarantee and use of anonymity, but a smaller target group could mean that the results can still be easily tracked back to individual participants.
+ The Data Protection Act (DPA) makes confidentiality a legal right
+ Recording of personal data is only acceptable if the data is not available in a form that could identify participants.
What does the ethical issue of privacy involve?
2 answers
+ Can be difficult to not invade privacy when studying participants without their awareness, i.e. in a field experiment
+Normal privacy should ideally be withheld (like having privacy in your own home) t maintain mundane realism.
How can you help to deal with issues of Informed Consent in an experiment?
3 answers
+Obtaining formal consent to the experiment and conformation of knowing the nature of, purpose of and their role in the experiment.
+ Could ask presumptive consent
+ Offering and exposing the right to withdraw
What is presumptive consent?
When a researcher asks a group of people who are similar to a participant if they would take part in this study, after revealing the aims and conditions fully. If the answer is yes, it is presumed the participant would give consent. This is useful if the study requires deception to obtain the required results or reduce demand characteristics.
How can a researcher deal with the ethical issue of deception in a study?
+ The need for deception should be reviewed and approved by an ethics committee
+ Participants should be fully debriefed after the study, which involves revealing the studies true nature and allowing the participant to discuss any concerns this raises.
+ Participants should also be offered the option to withhold their data from the study
How can a researcher factor in the right to withdraw in their design of a study?
+ Participants should be informed of this right at the start of the study
How can researchers improve protection from harm within a study?
2 answers
+ Avoid any risks greater than those experienced in everyday life
+ Stop the study if harm is suspected
How can researchers factor in a participants privacy in an experimental study?
name one limitation to this
+ Not studying anyone without their consent unless it is public behavior performed in a public place (i.e. PDA in a park)
+ However, there is no universal agreement of what makes a public place
What is meta-analysis?
A researcher looking at all the findings and results from similar studies investigating similar aims and producing a statistic to represent the overall effect/result (i.e. saying a specific weight-loss programme causes on average 30kg weight loss, based on the effect in multiple individuals)
What is one strength of meta-analysis?
Can increase population validity, as the result is from a larger group of participants.
What is an unstructured observation like?
Include example
When the researcher records ALL relevant information, without much specification or system. This can be problematic, as it can lead to the researcher only recording easily visible or eye-catching information. These may not be the most necessary or relevant behaviors. An example o this might be recording everything a person with anxiety disorder disorder does in a day, in order to try and analyse anxiety disorders.
What is a structured observation like?
It is considered the preferable method of observation, as it uses various systems to organise observations. Two main ways to structure observations are to use Behavioral Categories and Sampling Procedures.
What are Behavioral Categories and how are they useful?
helps us to compartmentalise actions into small, specific behaviors. For example, when recording a baby, instead of just observing an entire day with them, observing categories such as smiling, crying, sleeping, making fists, drooling, screaming etc. Behavioral categories should be objective and explicit enough that the observer should not have to make inferences from it, cover all possible behaviors in that situation/action and be mutually exclusive, meaning only one occurs at a time.
What are sampling Behaviors?
What are 2 examples?
Used in an unstructured observation, when there is too much information to record every piece of continuous data observed.
Event sampling- counting the number of times a target action or behavior occurs in an individual or group.
Time Sampling- Recording behaviors in a given time frame (i.e. level 2 obvs at Highfield)
What is the difference between PET and fMRI scans?
PET scans give more information as it has colours. It involves putting radioactive dyes in the body, that allow different tissues to show up on camera. It costs around £2000 per brain scan.
fMRI scans are similar, but only cost £500 per scan , and are less informative.
How are Correlations useful in psychology?
Used to analyse two co-variables and their relationship with each other, used when it would be unethical to do an experiment and it would be impractical and/or impossible to control the variables.
What is the correlation coefficient?
A measurement of the strength and the direction of a correlation.
What is the figure for a Perfect negative correlation Coefficient?
-1.0
What is the figure for a Negative correlation Coefficient?
-0.5
What is the figure for no correlation Coefficient? (no correlation or relation)
0.0
What is the figure for a Positive correlation Coefficient?
+0.5
What is the figure for a Perfect correlation Coefficient?
+1.0
What are two examples of direct investigator effects?
+ when the researcher uses extraneous or confounding variable that change the behaviors/DV of the participants
+ If, during observation or an interview, an investigator shows enthusiasm for specific enthusiasm or delivers questions with varying enthusiasm
What are 3 examples of Indirect Investigator Effects?
+ operationalizing variables in a way that makes the desired result more likely
+ limiting the duration of the study to make it easier to get the determined result
+ Keeping the standardised procedure not specific, leaving more room for the investigator to influence the results.
Name two ways to reduce investigator effects
+The single blind design- Keeping the participant unaware of research aims
+ The double blind design- both participant and investigator are kept unaware of aims.
Name the three types of experimental design
Repeated measures design, independent group design and matched pairs design
Describe independent group design
Participants allocated into two or more groups, one condition/IV tested on each group. Allocation usually done using random technique.
Describe matched pairs design
Participants are paired up based on matching or similar key variables. One person from each pair then completes the method with a different IV/condition.
Describe Repeated Measured Design
Each participant takes part in each condition sequentially.
What is one validity issue that could affect a systematic sample
Population clusters
What are three factors one should consider when doing content analysis?
+ Sampling method
+ Coding the data
+ Method of representing data
What does sampling method mean in content analysis?
The researcher must decide how to choose the sample they are analysing (e.g. if they are analysing a book, do they analyse every page or every five pages? or if analysing TV, do they note behavior every 30 seconds or just whenever specific behaviors occur?)
What is content analysis?
A kind of observational study in which behavior is usually observed indirectly in visual, verbal or written material.
What does coding the data mean in content analysis?
Coding in content analysis is often the different behavioral categories.
What are the two ways of representing data in content analysis?
Via quantitative or qualitative data.
What is one strength of content analysis relating to validity?
It has high ecological validity because it is an analysis of accounts of people’s real behavior.
What is one strength of content analysis relating to reliability?
When the source being analysed is public/can be accessed by others, others can replicate the content analysis and therefore the observations from the original content analysis can be tested for reliability.
What is one limitation of content analysis relating to bias?
Observer bias can reduce the objectivity and validity of findings as different observers may interpret the behavioral categories differently.
What is observer bias?
A for of bias where the researcher incorrectly ascertains or records data from a study due to knowing what they are analysing, so therefore losing objectivity.
What is one limitation of content analysis relating to culture?
Content Analysis is likely to be culturally biased as the verbal, visual or written content will be determined by the language and culture of the observer, so the source material and behavioral categories can be influenced by this.
What is the aim of an experiment?
a statement of what the researcher intends to find out in a research study
What is an experiment?
A research method where causal conclusions can be drawn because an independent variable has been deliberately manipulated to observe the causal effects on the dependent variable.
What is a hypothesis?
a precise and testable statement about the assumed relationship between variables.
How do we operationalise?
Ensuring that variables are in a form that can be easily tested. A concept such as “educational attainment” needs to be specified more clearly if we are going to investigate it.
What are standardised procedures?
A set of procedures that are the same for all participants in order to be able to repeat the study. This includes standardised instructions.
What is the flow chart for the scientific method of experimentation?
observe →→ state expectations →→ design a study →→ see if your expectations were correct
What is the difference between aims and a hypothesis?
Aims are an initial statement of what the researcher plans to investigate whereas the hypothesis is a formal and testable statement of the relationship between variables.
When will psychologist use a directional hypothesis?
When past research suggests that the findings will go in a particular direction.
When will psychologist use a non-directional hypothesis?
When there is no past research supporting the hypothesis or when past research is contradictory to the hypothesis.
What is a pilot study?
A small- scale trial run of a study to test any aspects of the design, with a view to making improvements. This helps researchers not waste time and money on an ineffective large-scale study.
What is order effect and in what experimental design does this occur?
How can this be dealt with?
The order of the conditions that are carried out in a matched pairs design affecting performance.
The main way that order effects are dealt with is using counterbalancing.
What is counterbalancing?
An experimental technique used to overcome order effect when using a repeated measures design. Counterbalancing ensures that each condition is tested first or second in equal amounts.
What are the two ways of counterbalancing?
AB or BA- Divide participants into two groups- in group 1, each participant does A and then B. In group 2 each participant does B and then A. This is still repeated measures as they are still each doing both conditions.
ABBA- All participants take part in each conditions twice, i.e. condition a and then condition b in trial one and later the condition b and condition a.
What is experimental realism?
If the researcher makes an experimental task sufficiently engaging the participant pays attention to the task and not the fact that they are being observed.
What are participant variables?
They are extraneous variables that is based on a characteristic of extraneous variables. An example of this could be age, intelligence, motivation etc.
What are situational variables?
They are types of extraneous variables that are a result o the research situation that may influence participants’ behavior. An example of this could be order effect.
what is controlled observation?
A form of investigation in which behavior is observed but under conditions where certain variables have been organised by the researcher.
What is inter-observer reliability?
The extent to which there is agreement between two or more observers involved in observations of behavior.
What is an intervening variable?
A variable that comes between two other variables, which is used to explain the association between those two variables.
What is a systematic review?
When a person looks through databases and journals and, after identifying the aim, searches for research studies that have addressed similar aims/hypothesis.
What is qualitative data?
Data that can’t be counted or quantified but it can be turned into quantitative data by placing the data into categories and then counting the frequency.
When do we use the sign test?
A test that is used when looking at paired or related data using nominal data.
How do we find the calculated value in the sign test?
We add up all the signs (the pluses and minuses separately) and select the smaller value. This smaller value is then S.
What is N in the sign test?
The total number of scores (ignoring any zero values)
What is peer review?
The assessment of scientific work by others who are experts in the same field. The intention of peer reviewing is to ensure that any research conducted and published is of high quality.
What does the Parliamentary Office of Science and Technology suggest are the three main purposes of peer review?
Allocation of Research Funding- research is paid for be various government and charitable bodies
Publication of research in academic journals and books- Scientific or academic journals provide scientist with the opportunity to share the results of their research.
Assessing the research rating of university departments
What is thematic analysis?
To identify qualitative data in media that can be analysed by identifying repeated themes.
What is the best way to assess reliability?
Having two or more observers make separate recordings and then compare these records. You can then quantify the agreement by calculating a correlation coefficient and then test the significance with a statistical test.
What are two ways of improving reliability?
Standardisation and reducing ambiguity
What are two ways of assessing reliability?
Test-retest reliability- seeing if a test or questionnaire gets the same result every time
Inter-interviewer reliability- assessing the reliability of one interviewer by comparing answers on one occasion with answers from another occasion with the same interviewer.
What is falsifiability?
It refers to being able to prove a hypothesis wrong. This is necessary for any study.
What is the idea of the importance of R in terms of statistical tests?
If the name of a statistical test has an R in it (Spearman’s, Pearson’s, Chi-Squared, unrelated and related t test) then the calculated value should be greater than the critical value. If there is no R, (Sign test, mann-whitney U and Wilcoxon T) then the critical value should be less than the calculated value.
What will the mean and median be like if the data shows a negative skew?
The mean and median will be less than the mode
If the mean is less than the median
How is the mode different than the mean and median?
The mode is different than the mean and median as it is a nominal measurement rather than a numerical measurement.
What does data where the mean. median and mode are all nearly the same indicate?
Normal distribution
What will the mean and median be like if the data shows a positive skew?
The mean/median will be higher than the mode
