Overview of Research Methodology Essentials

Title: Research methodology and methodological concepts *12 URL Source: blob://pdf/033f93bd-9912-4492-94af-9f5433c4926e Markdown Content: # Research Methodology Learning outcomes 1. 6 main research methods: experiments, self-reports, observation, case studies, correlations, and longitudinal studies 2. Methodological concepts, including aims, hypotheses and variables, experimental design, control of variables and sampling 3. Ethical guidelines for working with humans and non-humans animals 4. Ways of evaluating research: validity, reliability, and replicability 5. Types of data and data analysis Table of contents 1. Experiments 2. Experimental designs 3. Self-reports: Questionnaires 4. Self-report: Interviews 5. Randomised control trials 6. Case studies 7. Observations 8. Correlations 9. Longitudinal studies What are experiments? Discover cause-and-effect relationships between variables Help reveal the causes of behavior, thoughts, and feelings -> compare how people behave in different situations ( conditions ), changing one thing at a time ( the independent variable ) and observing how this affects another thing ( the dependent variable ) How to be sure whether changes in the dependent variable were really caused by the independent variable? Think about other variables that might affect DV Make sure they are controlled Standardised procedure/standardised instructions # Experiments (1/4) Does this mean wearing masks cause people to get covid? Discuss "False cause fallacy" Occur when someone mistakenly assumes that one event caused another simply because they occurred in sequence, or that one event is the direct cause of another without sufficient evidence A study shows that people who drink red wine tend to live longer The risks of incorrect assumptions Main features of experiments: Compare data collected from the experimental group with data collected in the control group to check for cause-and-effect IV manipulated; DV measured Cause-and-effect established only if all other variables are controlled Standardised procedure and instructions: participants have the same experience; study can be replicated # Experiments (2/4) Laboratory experiments Location : Laboratory> setting (location) at which experiments takes place Controls : Controlled variables kept same between all levels of IV # Experiments: types Laboratory experiments Evaluation Validity : high levels of control enables researchers to establish a cause-and-effect relationship Low ecological validity Validity affected by Demand Characteristics Reliability : high standardized; can be replicated Ethics : should be more ethical # Experiments (3/4) Field experiments Location : Real world setting Controls : difficult to maintain control over variables that could affect DV # Experiments types Field experiments Evaluation Validity : High ecological validity; Less likely to show Demand Characteristics Reliability : much less control in a field; difficult to replicate due to limited ability to standardise the procedure Ethics : no consent, no right to withdraw # Experiments (4/4) Independent measures design Participants put into 2 groups (experimental and control group) Data from control group provide baseline to compare data from experimental group Random allocation and participant variables Randomly allocated to groups Without random allocation members of one group might share common characteristics (participants variables) # Experimental designs (1/2) Repeated measures Same group allocated to both experimental and control group (acts as their own control group) Problems: demand characteristics and Order effects. Demand characteristics Cues in set-up of experiments which might give away clues about aim and hypothesis Order effects When participants perform task twice in different conditions, behavior in second might differ already performed task once, so perform better the second time # Experimental designs (2/2) The placebo effect To isolate how much of any improvement is due to the drug, and how much is due to expectancy effects Placebo treatments used Single-blind : participants do not know whether they received real treatment or placebo RCT used for psychological treatment such as Cognitive Behavioral therapy (CBT) Double-blind design Neither researcher nor participants know which is treatment which is placebo # Randomised control trials (1/2) Evaluation Validity : double-blind design reduces experimenter bias; reduces demand characteristics Reliability : double-blind allows researchers to test interventions (eg: treatment/therapies) Ethics :Valid informed consent is not possible Participants randomly assigned to control do not have access to treatment that could significantly improve health # Randomised control trials (2/2) Questionnaires Printed, completed by hand, distributed digitally, completed online # Self-reports: Questionnaires (1/2) Closed questions Offer a fixed choice of answers/ uses scale such as a Likert scale. Open questions Allow participants to express opinion freely/ Question starts with why when/ Allows researcher to understand complexity of issue Self-reports: Questionnaires (2/2) > Evaluating questionnaires Interviews Researcher directly speaks to individual participants, ask questions and records responses. Generally done with small number of participants Time consuming # Self-reports: Interviews(1/3) Types of interviews # Self-report: Interviews (2/3) Identify the structured and semi-structured interview script Evaluating interviews # Self-report: Interviews (3/3) Advantages Disadvantages Gather more in-depth data Participants more likely to give socially desirable answers Structured and unstructured interviews allows to gain good insight to area being investigated People might feel uncomfortable to be interviewed face to face, hence prefer telephone or video calls Allows to gather information about non-verbal communication (eg: posture, facial expressions) Some might not feel comfortable talking over the telephone/ researcher cannot see the participants hence cannot use non-verbal signals to help them engage with participants Telephone interviews are less time consuming, more cost-effective Detailed investigations of individual people or small groups (units) of people; from same family, single shop, health clinic Tend to study unusual and rare cases Key feature: data gathered using a number of different techniques (interviews, observations, psychometric tests Triangulation ) Data often gathered over a long period of time longitudinal studies Data gathered maybe qualitative or quantitative data When psychologists write for publications, they often begin with a case history # Case studies (1/2) Advantages Disadvantages Gather rich detailed information about individual or small group of individuals Researcher could become less objective when analysing data since they may be influenced by how they feel about the participant, reducing validity Case studies use more than one method of gathering information , researchers can triangulate the data to ensure it validity Hard to replicate case studies simply because it is a detailed analysis one one unique individual or a small group of individuals # Case studies (2/2) Evaluating case studies Involves watching peoples behavior and recording it May observe either human or non-human animal behavior # Observation (1/5) Naturalistic : observes participants in uncontrolled, real-world setting Controlled : laboratory settings, standardise the situation to every participant Structured : predetermined, limited no.of behaviors using a checklist Unstructured : record all behaviors that are relevant, researcher writes notes Overt and covert Participant and non-participant Structured and unstructured Naturalistic and controlled Overt : participants know Covert : participants do not know Participant observation : researcher joins group Non -participant observation : researcher observes from a distance # 01 02 # 03 04 # Observation (2/5) Evaluating observation # Observation (3/5) Advantages Disadvantages Structured observations are more reliable than unstructured observations as the behaviours that are to be recorded are decided in advance and operationalised. Behaviour that is being observed will be broken down in components that can be easily identified and recorded. This ensures that different observers are more likely to record data consistently . Structured observations can also be subject to observer bias as they rely on the observer's own judgement as to what behaviour to record and are therefore subjective. Unstructured observations can be used as part of a pilot study to give researchers a good understanding and overview of the range of behaviours that they might observe. The researchers would then choose a limited number of these to observe for their main research. In unstructured observations, it can be difficult to record all behaviour and some of the subtler (but interesting) behaviours might be missed. Evaluating observation Advantages Disadvantages An advantage of overt observations is that it is more ethical because participants know they are being observed so it avoids having to deceive participants. A disadvantage of overt observations is that participants may show demand characteristics because they know they are being watched and therefore may not act in a natural way An advantage of covert observations is that it avoids any potential demand characteristics because the participants do not know they are being observed. A disadvantage of covert observations is that they are less ethical than overt observations because participants do not know they are being observed and studied. Evaluating observation Advantages Disadvantages An advantage of non-participant observations is that the observer is likely to be more objective in their observations as they are not personally getting involved in the study. A disadvantage of non-participant observations is that because the observer is watching from a distance some behaviours may be missed. An advantage of participant observations is that the observer is not viewing from a distance and therefore may gain a greater understanding of the participants' behaviour. A disadvantage of participant observations is that the observer may become too involved with the people that they are observing, and become less objective in their observations Evaluating observation Advantages Disadvantages An advantage of naturalistic observations is that behaviour is likely to be normal as the participants are in their own natural settings and the researchers do not interfere in any way. Therefore it is likely to be high in ecological validity. A disadvantage of naturalistic observations is that it is much harder to control for variables that might affect the participants' behaviour. An advantage of controlled observations is that they can be more easily replicated by other researchers as they can use the same behaviour schedule and it is easier to standardise the situation for all participants. This makes controlled observations more reliable. A disadvantage of controlled observations is that behaviour may be less natural if participants are aware that they are in a controlled setting. Researchers might also miss key behaviours if they have a very rigid behaviour schedule. Looks at the relationship between 2 variables These are called co-variables For each participant, researchers will gather 2 sets of data which they can plot on a scatter graph to see whether there is a correlation between 2 measures/variables # Correlations (1/8) Operational definition: Operationalizing the variables means to state exactly what they are and how they will be measured Example: Co-variable 1: average time spent on playing video games per week measures in hours Co-variable 2: reaction speed on an app measured in minutes, seconds, and milliseconds It is ALWAYS important to operationalize hypothesis for correlations # Correlations (2/8) Operational definition: Correlations are often used by researchers: To investigate new areas of psychology before conducting experimental research, OR in cases where it is not practical or ethical to manipulate variables Various methods used for data collection, AS LONG AS THE DATA IS LONGITUDINAL Data is analyzed by plotting pairs of scores in one another on a scatter graph (to see whether there is any correlation or not) # Correlations (3/8) Positive, negative, and no correlations Positive correlation 2 variables INCREASE together EXAMPLE : Dement and Kleitman (sleep and dreams) - the longer the duration of REM sleep (in minutes) the more words participants used to describe their dream. # Correlations (4/8) Negative correlation When one variable INCREASES, the other variable DECREASES EXAMPLE : In Baron-Cohen et al. (eyes test), the higher the Eyes Test score the lower the Autism Spectrum Quotient score # Correlations (5/8) No correlation no consistent relationship between scores EXAMPLE : No correlation between time spent on mindfulness activities between the weekly MBSR sessions and grey matter concentration. # Correlations (6/8) Strength of the relationship between 2 variables can be represented by a CORRELATION COEFFICIENT . In general, the closer to 1 (for a positive correlation) OR the closer to -1 (for a negative correlation), the STRONGER the correlation is. IS CORRELATION COEFFICIENT VALUE OF +0.75/-0.75 A STRONG COEFFICIENT? # Correlations (7/8) Evaluating correlations # Correlations (8/8) Advantages Disadvantages Correlations allow psychologists to investigate new areas of research to see whether they are worth investigating further with experimental methods . Correlations can only show the strength of the relationship between two variables and cause and effect cannot be inferred. They also do not tell us why that relationship has occurred Correlations allow researchers to investigate areas where it is not practical or ethical to manipulate variables. The presence of a third variable cannot be ruled out in correlations. A third variable that is not being measured may have influenced the relationship . Importance of objective data collection Involves following the same group of individuals over an extended period of time and giving them tests or tasks at various intervals during that time to see how they change and develop in their thoughts, feelings and/or behaviour. Participants of longitudinal designs need to be re-contacted over time for repeated testing. This testing needs to be standardised for valid and reliable comparisons to be made between participants' scores over time. # Longitudinal studies (1/3) Longitudinal studies use a variety of techniques to gather data , from psychometric testing to interviews and questionnaires. Experiments can also use a longitudinal design where the experimental group is given an intervention (such as a wellbeing programme, an educational intervention or a dietary supplement) and compared over time to a control group that is not given the intervention. Follow-up sessions allow psychologists to measure the long-term impact of treatments and interventions on the participants. # Longitudinal studies (2/3) Longitudinal studies (3/3) Evaluating correlations Advantages Disadvantages Because longitudinal studies are following the same group of participants as they change and develop over time, researchers do not have the issue of participant variables and participant effects Due to the long-term nature of the research, it is more prone to attrition where participants drop out of the research, either because they move away and lose contact with the researchers, they decide they no longer wish to take part or their life circumstances change. Allow some topics to be studied which might not be possible using other research methods, such as child development over time. By the time longitudinal research is completed, it may no longer be generalisable due to changes in society This means it may lack temporal validity Group work Make your own experiment for one of the following topics 1. Social anxiety and baron-cohens eye test for social sensitivity 2. Differences in intelligence between millennials genZ 3. Bullying and depression 4. Childhood trauma and adult depression Your explanation should include 1. Type of research design (correlational/ experimental/ case study ect) 2. Type of data (quantitative/qualitative) 3. Procedure 4. Type of instruments used (interview/ questionnaire) 5. The results you expect 6. Weaknesses and strengths (ethics included) Aims The aim is a statement that explains why the researcher is doing the research and what they hope to achieve. Hypotheses A hypothesis is a statement where the researcher predicts what they think will happen in the research. Hypothesis is operationalised; meaning that the independent and dependent variables (in an experiment) or the co-variables (in a correlation) are clearly defined. The units of measurement (e.g. milliseconds) should always be included. For example, a phobia could be measured using a score from a self-report questionnaire, heart rate from a heart rate monitor (in beats per minute), a participant's body language from an observation (e.g. number of times they touch their face or hair, cross and uncross their legs, etc.) or a measurement of cortisol (a stress hormone) in their saliva. Directional hypotheses With a directional hypothesis: The researcher will predict whether the experimental group will perform significantly better or worse than the control group (or in a correlation, whether it will be a positive or a negative correlation). Directional hypotheses tend to be used where previous research has been done in the area, so the researcher has a good idea of what results to expect Example: People with childhood trauma will develop depression in adulthood People who drink coffee will have quicker reaction times than people who drink a cup of water Non-directional hypotheses For a non-directional hypothesis: Researcher states that there will be a significant difference between the experimental and control groups but does not state in which direction that difference will be .For example, whether the experimental group will be faster or slower. Similarly, with a correlation, the researcher will state that there will be a significant correlation between two variables, but not whether it will be a positive or negative correlation . Non-directional hypotheses Researchers tend to use non-directional hypotheses: When there has not been previous research conducted in the area that they are researching When the research is contradictory regarding what the possible outcome might be Null hypothesis The null hypothesis is the statement that the independent variable will have no effect on the dependent variable. OR That two variables are not correlated with each other. It is important for the null hypothesis to also be operationalised. Null hypotheses normally start with 'There will be no difference/correlation ... ', and they usually finish by saying that 'any difference that does arise will be due to chance' Data analysis Psychologists analyse quantitative data using descriptive statistics These help to summarise the data and make it easier to see whether there are any obvious patterns or trends. Psychologists use a branch of mathematics called inferential statistics, which allows them to calculate the probability (chance ) that their result would have arisen if the null hypothesis was true. This is called a p value. Having completed their analysis, the researchers will decide whether to accept or reject their null hypothesis. The null hypothesis is rejected if the p value (the probability of a significant result having occurred due to chance) is less than 0.05 (1 in 20) and it is accepted if it is more than 0.05 (1 in 20). Controlling of variables When conducting experiments, researchers attempt to control all variables other than the one being manipulated (the independent variable) Controlled variables are 'held constant' , or standardised , between the two or more conditions This allows the researcher to state that any changes in the dependent variable were caused by the independent variable . Standardisation of a procedure In order to make research replicable so that it can be checked for reliability, it is essential that the procedure is standardised for all participants. In an experiment, the only difference should be what the researcher is manipulating (the independent variable). However, all research, whether it is experimental or not, should follow standardised procedures. This includes participants all receiving the same information and instructions (unless this is something that is being manipulated as part of the independent variable) . Participant variable Participant variables are aspects of a person's background , personality , cognitive abilities , health and wellbeing , etc., that affect how they behave or respond in a study. Individual differences between people are of course inevitable and this is why psychologists work with averages calculated from larger groups of people. However, participant variables can cause problems in experimental designs if some aspect of the participants (e.g. empathy, aggression or intelligence) varies systematically between the two or more experimental and control groups. Participant variable For example; if all passengers in Piliavin et al.'s 'cane' condition were high in empathy and all the people in the drunk condition were low in empathy, it would be impossible to know whether aspects relating to the victim (drunk or cane) affected helping behaviour, or whether it was individual differences (high and low empathy) between the passengers that affected behaviour. How to overcome participant variables? 1. Random allocation of participants to the experimental and control groups can help overcome the effects of participant variables on the results. This was not possible in Piliavin et al.'s study as it was a field experiment But In Andrade's study (doodling) participants were randomly assigned to the experimental (doodling) and control (non-doodling) groups to ensure that people with naturally good memories did not all end up in the doodling group How to overcome participant variables? 2. Using a matched-pairs design can help reduce the impact of participant variables, as researchers can match participants in both the experimental and the control groups on key factors that could impact on the research (age, handedness, cultural background or socioeconomic status) For example, in Bandura et al. (aggression) participants were matched on their existing aggression levels, so that each group (experimental and control groups) contained children with lower and higher levels of aggression. Again, if the most aggressive children had all ended up in one of the experimental groups, it would be unclear whether they bashed Bobo due to observing an aggressive adult or they would have done this on their own anyway. Situational variables Situational variables refer to any aspect of the environment that could impact participants' behaviour and affect the results. Environmental variables include factors like: Temperature of the room Lighting and noise levels It is important for researchers to control these variables as much as possible so that each participant experiences the same environment, AND it is ONLY the independent variable that is affecting the dependent variable. It also makes the research replicable Uncontrolled variables These are factors that are not being measured or controlled by the researcher and can have an unwanted effect on the dependent variable. These could be things such as the temperature of the room or noise levels. Types of data Quantitative data refers to any data that is numerical . For example, closed questions on questionnaires produce quantitative data Qualitative data refers to non-numerical data, usually in words but sometimes in images (e.g. photographs). Open questions produce qualitative data as participants are able to freely express their thoughts and feelings. Types of data Strengths Weaknesses Qualitative data Qualitative data provides researchers with a more in-depth understanding of what their participants are thinking and feeling and the reasons behind their behaviour, which may give the data greater validity Qualitative data analysis may be more at risk of researcher bias as the interpretation of the data is more subjective . Therefore, the findings can be less reliable. Quantitative data Quantitative data can be more easily compared. The analysis of quantitative data is objective and so less prone to researcher bias. Quantitative data does not always allow us to fully understand what a participant is thinking or feeling . Some scales may limit how a participant can respond and they may feel that none of the responses reflects how they feel, making the data less valid. Types of data Subjective and objective data Subjective data means that the data can be influenced by a person's personal thoughts, feelings or opinions. This is more likely to be an issue with qualitative data, which requires interpretation by the researcher when being analysed. Objective data is unbiased, factual and not influenced by a person's personal thoughts or opinions. For example, research that uses ratings scales that are quantifiable is more objective as the data does not require interpretation by the researcher. In scientific research, it is important to be as objective as possible. Researchers can improve objectivity by getting another researcher, who does not know the aim of the study, to interpret the data. This person is likely to be more objective than the primary researcher, who will have already formed hypotheses about what the data will show. Types of data How to improve objectivity? Researchers can improve objectivity by getting another researcher , who does not know the aim of the study, to interpret the data. This person is likely to be more objective than the primary researcher, who will have already formed hypotheses about what the data will show. Sample and population Sampling refers to how researchers obtain participants for their research. Researchers will have a target population of people to whom they wish to generalise their findings. For example, the target population might be the global community of Cambridge International A Level psychology students. It would be practically impossible to test the entire population and instead the researcher selects a sample from the target population. Collectively, researchers refer to their participants as the sample .Sampling of participants Sampling techniques 1. Opportunity sampling: There are a number of ways that researchers can select their sample. Opportunity sampling is a technique in which researchers select participants who are readily available. For example, a psychologist might know six teachers from different countries around the world who teach Cambridge International A Level. The psychologist will ask those teachers if they can use their students. Many research studies are conducted by university lecturers using their own students. Sampling of participants 2. Random sampling: With random sampling, every person in the target population has an equal chance of being chosen to participate in the research. For example, if the target population is Cambridge International A Level students, this means that each Cambridge student across the world would have an equal chance of being selected as a member of the sample. This would involve getting the names of all of the students and putting them into a random name generator (computer program), which would pick a sample of 100 students, for example. Sampling of participants 3. Volunteer (self-selecting) sampling: participants put themselves forward to take part in a piece of research. Researchers may advertise for participants in many ways including posters, flyers, direct mail, e-shots, newspapers, radio and online forums. For example, if the researchers were looking to study a condition such as prosopagnosia (a condition where people struggle to recognise faces), they could find a Facebook group of people with the condition and advertise their research in the group so that members can volunteer (self-select) themselves to take part. Sampling participants Strengths Weaknesses Opportunity sampling Quick and easy as researchers use people who are readily available at the time of the research. May result in a larger sample and likely to mean research can be conducted without delays. Often unrepresentative of the target population . Findings may not be generalisable. Volunteer sampling Participants have volunteered to take part so they are motivated and willing -drop-out rates are likely to be low . This is important in longitudinal research where attrition may significantly affect validity May not be representative as only certain personality types volunteer to take part in research , such as those with more pro-social attitudes, those who are more motivated than the rest of the population and those who have a personal interest in the research topic. Sampling participants Strengths Weaknesses Random sampling Should be representative of the target population , making the findings more generalisable (if the sample is large enough). Time-consuming and expensive . Potential participants may not wish to take part in the study . All potential participants must be contacted and arrangements must be made for them to visit the location where the research is being he Ethics and human participants Ethics Ethics are critical component of all psychological research Wherever in the world there is psychology, there are always ethical guidelines Informed consent and right to withdraw- core principles of ethics Ethical guidelines: Serve to protect participants Ensure they do not come to any hard Ensure that research that benefit wider society can be pursued Valid consent: This means participants know exactly what will be happening during the research before they agree to take part Consent must be provided before the data collection begins BUT is not considered valid until it is INFORMED As in, participants must know what they have signed up for and should be given a detailed, age-appropriate information Researchers must be able to provide evidence that consent was obtained Valid consent: Incase of some vulnerable groups, parents family members, or caregivers should also give consent Eg: consent from parents or caregivers when working with adults with low cognitive and communication abilities Parental consent must be obtained when working with children (under 16 or under 18 depending from country to countrys age restrictions) RIght to withdraw: Although participants give their consent, they must also be aware that they can withdraw from the research at any time without any pressure They must be informed that their data will be deleted or destroyed Researchers working with very young children or adults with significantly impaired communication ability should look for signals (verbal and non-verbal) that participants is no longer to participate Minimizing harm and maximizing benefit: Must aim to minimise harm to participants psychological well-being, personal values, privacy or dignity and mental health Risk of harm should be no greater than the participant might expect to experience in their everyday lives Should aim to maximise the benefits of their research at all stages of the research process Lack of deception: By obtaining informed consent, participants will be very clear about what they will be doing and the aim and purpose of the study Lack of deception means that the participants will not be deceived about anything about the research In reality, it will be difficult to conduct research without any deception as simply knowing the true aim of the research could change participants behavior When deception is involved, full debriefing is essential confidentiality: Participants results should remain confidential before and after publication of their data Particularly important when the data obtained is qualitative; more likely to be personal and thus more identifiable Data should be kept confidential while conducting the research and analysing the data Participants names should not be included with their data Privacy: In the context of observational research, participants should only be observed in public situations where they might be expect to be observed by others Not acceptable to observe people in their own homes without consent Debriefing: Happens after the participants has completed the research Purpose of debriefing: to explain any deception that has taken place to answer any questions the participants may have To ensure they leave the research in the same state in which they entered in (negative state of mind t should be turned to positive state of mind) Ethics and non-human participants Minimizing harm and maximizing benefit: Must seek to minimize harm, discomfort, and suffering to the animals Maximize the benefit of the research in applying their findings to helping other animals or animals of the same species Replacement: The use of animals is replaced with video footage from prior research or computer simulations Particularly important if animals are to be used for training or teaching purposes Species: Must important to use species or animals that are both scientifically and ethically suitable for the research Researcher: must have good understanding of the animals natural history and its ability to feel Researchers should know whether animals have been bred in captivity or not Researchers must choose the species that is likely to suffer the least Numbers: Must use the smallest number of animals possible to meet the research aim Use of pilot studies with small number to ensure that design is appropriate and testing what the researchers want to investigate before using a large sample of animals Good experimental design ensures the minimum use of animals Procedures- Pain and Distress: Should avoid causing death, diseases, psychological or physical discomfort to animals Should look at enriching animals environment rather than depriving Should consider balance between harm caused to the animals and the benefits that would be gained Pain is ONLY ETHICALLY JUSTIFIABLE: when there is no alternative research method available research would have significant scientific or educational value. Surgical procedures should be conducted using anaesthesia to reduce pain and prevent infection If termination is needed, must be done swiftly Housing: Social and natural behavior of the species should be considered when housing the animals Isolating animals who would normally live in social groups could cause stress Overcrowding could also cause stress and aggression Using same cages could reduce negative impact Reward, deprivation, and aversive stimuli: If researcher is considering use of food deprivation, they must have a good understanding of the species normal eating and drinking habit to ENSURE THAT DEPRIVATION IS KEPT TO A MINIMUM. Validity Validity Extent to which the researcher is measuring what they think they are measuring and the extent to which the findings are useful and meaningful Historical example of a psychometric test that lacked validity > Army Alpha and Beta IQ tests: The term 'IQ' stands for ' intelligence quotient'. In 1917, a psychologist called Yerkes worked with colleagues to create IQ tests for army recruits in the United States of America, to see which section of the army would be most appropriate for them, based on their IQ levels. Many of the recruits were immigrants who had not lived in the country very long. However, the 'IQ' tests contained questions such as 'What is Crisco?' and 'Who is Christy Mathewson?'. These questions were impossible to answer unless the recruit had a good knowledge of American culture. This led to recruits from certain nations experiencing prejudice as they (understandably) received low IQ scores. This is a good example of a test that lacks validity - it was testing cultural knowledge, not IQ . Ecological validity The extent to which behaviour that participants demonstrate during research relates to how they would behave in their real lives. Research conducted in normal, real-life settings , as in field experiments, is likely to be higher in ecological validity .However, some laboratory experiments can also be high in ecological validity if the setting is unlikely to impact on behaviour. Ecological validity > If research lacks ecological validity , then it cannot be generalised to real-life situations. However, where it is important for research to be conducted under controlled conditions, elements of the procedure can be made more ecologically valid. For example, if participants had to demonstrate their driving skills in a laboratory setting, it would have greater ecological validity if a participant were able to use virtual reality headsets with a steering wheel, brake and accelerator rather than using a computer keypad as controls. Subjectivity/Objectivity If a researcher is required to interpret behaviour, or the data collected, in any way, there is always a risk of subjectivity . This is where their personal thoughts , feelings or opinions may affect the validity of the findings. An example: researcher misinterprets a certain behaviour and records it in a way that was not meant. This would lead to the data not being valid. In terms of data analysis, qualitative data is most at risk of subjectivity as it requires a researcher to analyse the meaning behind the data and draw conclusions based on this. # Evaluating studies based on their validity Subjectivity/Objectivity > The researcher's own interpretation may be influenced by their personal biases , either consciously or > unconsciously , leading to low validity. For example, Milgram (obedience) needed to interpret the behaviours of the participants as they delivered what they believed to be electric shocks to another person and decide what data to record. Having previously conducted a pilot study, he may have already had certain expectations about what behaviours to expect. Subjectivity/Objectivity > If data collection or analysis is not influenced by a researcher's personal thoughts, feelings or opinions , then it is said to be objective . Scientific equipment - for example, brain scanning techniques such as EEGs used in Dement and Kleitman (sleep and dreams) and MRis used in Holzel et al. (mindfulness and brain scans) - is only capable of recording what biologically occurs, thus this method of collecting data is completely objective. Subjectivity/Objectivity It is worth noting , however, that the images produced by brain scans still require the subjective interpretation of researchers . Quantitative data is also more objective than qualitative as it can be statistically analysed. The less risk the data has of being influenced by another person's interpretation, the more objective and some would argue, the more valid. Demand characteristics Participants might try to guess the aim of the research and change their behaviour based on what they think the researcher is looking for. This would reduce the validity of the study. Any features of the research environment which may give away the study's aim are called demand characteristics. Demand characteristics This is why some studies use deception to avoid the participants' behaviour being affected. In Milgram (obedience), participants were deceived in a variety of ways to attempt to hide the study's true purpose. One of the strengths of using animals as participants is that they do not actively try to work out what the study is about . Therefore, demand characteristics are reduced and validity increased. This said, animals may unconsciously pick up on subtle cues from researchers, which may affect their behaviour. Generalisability Findings should only be generalised to the target population IF the sample was sufficiently representative. This will depend upon the sampling technique, for example random samples are typically more representative than opportunity or volunteer samples. The extent to which findings can be generalised beyond the sample is sometimes called 'population validity' .For example, the study by Piliavin et al. (subway Samaritans) looked at helping behaviour on a New York subway carriage. It may not be appropriate to generalise their findings to all cultures because helping behaviour may differ in other countries, based on cultural values. Generalisability > Whether the results from a study can be generalised to real-life situations or not is called ecological validity. For example, as the study by Piliavin et al. (subway Samaritans) took place on a normal subway journey, as part of people's real lives, it is possible to generalise the findings to similar real-life contexts Temporal validity > Refers to whether the results of a study can be generalised to a different time period. For example, Bandura's study was conducted in the 1960s and the results may not be the same today as modern children have access to personal technology such as phone and tablets and gender roles have also changed significantly. This means the findings may be era-bound and said to lack temporal validity . Other studies, such as Hassett's study of sex differences and play in monkeys, are less likely to be affected by temporal validity since they are based on a generally stable biological functions. Social psychological research may be most susceptible to issues with temporal validity, as cultural and societal norms change over time . Reliability Reliability is about consistency and can refer to the results from each participant and the overall findings, based on the whole sample. If you tested your participants again and the results were the same, we could assume the results were reliable. If we ran the whole study again, with a different set of participants (a replication) , we could check the reliability of the findings. If we got the same results again , we could conclude that these findings were reliable. # Reliability and replicability Replicability Replication means that other psychologists can repeat the study exactly to see if they get consistent results. Therefore, the ability to replicate a study is critical as it helps researchers to demonstrate the reliability of their findings - that is, to show that their findings were not due to chance. Replicability relies heavily on the extent to which the study has a standardised procedure that other researchers can follow in their replications. # Reliability and replicability Standardization: Standardisation of the procedure and instructions provided to the participants is an important way of making a study replicable. It also means that all aspects of the research are consistent for every participant Standardization For example, in the study by Bandura et al. (aggression), the model always displayed the same sequence of behaviours, in the same order and for the same length of time. The toys were laid out in exactly the same way in each of the rooms and the children were always given the exact same standardised instructions. The observer also scored the children's aggressive and non-aggressive behaviour in the same way according to the categories on the behavioural checklist. This high degree of standardisation means Bandura's study is easy to replicate and, therefore, his findings could be checked for reliability. Ways to increase reliability Owari 1. Describe the ethical guidelines of 'species and 'replacement' in relation to animals in research, using any examples (6m) 2. Mirabel is observing two groups of gorillas in the wild, a large group of 65 and a small group of 20. She is recording the number of close social relationships each gorilla has with other gorillas in the group. Outline why Mirabel must have a strong knowledge of the 'species before being allowed to conduct this work (2m) 3. lchika is conducting a study on recall of items from a menu. A waiter reads out a list of nine dishes to diners in a restaurant. They are then asked which dish they would like to order. lchika records whether they choose from the first, middle or last three menu items. She allows half of the participants to doodle on a napkin while they are listening to the waiter and the other half are not allowed to doodle. Explain one reason that lchika's study has high validity and one reason why it has Low validity (4m) 4. Explain what is meant by the terms 'inter-rater reliability and 'inter-observer reliability using any examples (6)

Title: Research methodology and methodological concepts *12

URL Source: blob://pdf/033f93bd-9912-4492-94af-9f5433c4926e

Markdown Content:
# Research Methodology Learning outcomes

1. 6 main research methods: experiments, self-reports, observation, case studies, correlations, and longitudinal studies 2. Methodological concepts, including aims, hypotheses and variables, experimental design, control of variables and sampling 3. Ethical guidelines for working with humans and non-humans animals 4. Ways of evaluating research: validity, reliability, and replicability 5. Types of data and data analysis Table of contents

1. Experiments 2. Experimental designs 3. Self-reports: Questionnaires 4. Self-report: Interviews 5. Randomised control trials 6. Case studies 7. Observations 8. Correlations 9. Longitudinal studies What are experiments?  Discover cause-and-effect relationships between variables

Help reveal the causes of behavior, thoughts, and feelings -> compare how people behave in different situations ( conditions ), changing one thing at a time ( the independent variable ) and observing how this affects another thing ( the dependent variable )

How to be sure whether changes in the dependent variable were really caused by the independent variable?

Think about other variables that might affect DV

Make sure they are controlled

Standardised procedure/standardised instructions

# Experiments (1/4) Does this mean wearing masks cause people to get covid? Discuss "False cause fallacy" Occur when someone mistakenly assumes that one event caused another simply because they occurred in sequence, or that one event is the direct cause of another without sufficient evidence A study shows that people who drink red wine tend to live longer The risks of incorrect assumptions Main features of experiments:

Compare data collected from the experimental group with data collected in the control group to check for cause-and-effect

IV manipulated; DV measured

Cause-and-effect established only if all other variables are controlled

Standardised procedure and instructions: participants have the same experience; study can be replicated

# Experiments (2/4) Laboratory experiments Location : Laboratory> setting (location) at which experiments takes place

Controls : Controlled variables kept same between all levels of IV

# Experiments: types Laboratory experiments

Evaluation

Validity : high levels of control enables researchers to establish a cause-and-effect relationship Low ecological validity Validity affected by Demand Characteristics

Reliability : high standardized; can be replicated

Ethics : should be more ethical

# Experiments (3/4) Field experiments Location : Real world setting

Controls : difficult to maintain control over variables that could affect DV

# Experiments types Field experiments

Evaluation

Validity : High ecological validity; Less likely to show Demand Characteristics

Reliability : much less control in a field; difficult to replicate due to limited ability to standardise the procedure

Ethics : no consent, no right to withdraw

# Experiments (4/4) Independent measures design

Participants put into 2 groups (experimental and control group)

Data from control group  provide baseline to compare data from experimental group Random allocation and participant variables

Randomly allocated to groups

Without random allocation  members of one group might share common characteristics (participants variables)

# Experimental designs (1/2) Repeated measures

Same group allocated to both experimental and control group (acts as their own control group)

Problems: demand characteristics and Order effects. Demand characteristics Cues in set-up of experiments which might give away clues about aim and hypothesis Order effects When participants perform task twice in different conditions, behavior in second might differ  already performed task once, so perform better the second time

# Experimental designs (2/2) The placebo effect

To isolate how much of any improvement is due to the drug, and how much is due to expectancy effects  Placebo treatments used

Single-blind : participants do not know whether they received real treatment or placebo

RCT used for psychological treatment such as Cognitive Behavioral therapy (CBT) Double-blind design Neither researcher nor participants know which is treatment which is placebo

# Randomised control trials (1/2) Evaluation

Validity : double-blind design reduces experimenter bias; reduces demand characteristics

Reliability : double-blind allows researchers to test interventions (eg: treatment/therapies)

Ethics :Valid informed consent is not possible Participants randomly assigned to control do not have access to treatment that could significantly improve health

# Randomised control trials (2/2) Questionnaires

Printed, completed by hand, distributed digitally, completed online

# Self-reports: Questionnaires (1/2)

Closed questions

Offer a fixed choice of answers/ uses scale such as a Likert scale.

Open questions Allow participants to express opinion freely/ Question starts with why when/ Allows researcher to understand complexity of issue Self-reports: Questionnaires (2/2)

> Evaluating questionnaires

Interviews

Researcher directly speaks to individual participants, ask questions and records responses.

Generally done with small number of participants

Time consuming

# Self-reports: Interviews(1/3) Types of interviews

# Self-report: Interviews (2/3) Identify the structured and semi-structured interview script Evaluating interviews

# Self-report: Interviews (3/3)

Advantages Disadvantages

Gather more in-depth data Participants more likely to give socially desirable answers

Structured and unstructured interviews allows to gain good insight to area being investigated

People might feel uncomfortable to be interviewed face to face, hence prefer telephone or video calls

Allows to gather information about non-verbal communication (eg: posture, facial expressions)

Some might not feel comfortable talking over the telephone/ researcher cannot see the participants hence cannot use non-verbal signals to help them engage with participants

Telephone interviews are less time consuming, more cost-effective  Detailed investigations of individual people or small groups (units) of people; from same family, single shop, health clinic

Tend to study unusual and rare cases

Key feature: data gathered using a number of different techniques (interviews, observations, psychometric tests  Triangulation )

Data often gathered over a long period of time

longitudinal studies

Data gathered maybe qualitative or quantitative

data

When psychologists write for publications, they often begin with a case history

# Case studies (1/2) Advantages Disadvantages

Gather rich detailed information about individual or small group of individuals

Researcher could become less objective when analysing data since they may be influenced by how they feel about the participant, reducing validity

Case studies use more than one method of gathering information , researchers can

triangulate the data to ensure it validity

Hard to replicate case studies simply because it is a detailed analysis one one unique individual or a small group of individuals

# Case studies (2/2)

Evaluating case studies  Involves watching peoples behavior and recording it

May observe either human or non-human animal behavior

# Observation (1/5) Naturalistic : observes participants in uncontrolled, real-world setting

Controlled : laboratory settings, standardise the situation to every participant

Structured : predetermined, limited no.of behaviors using a checklist

Unstructured : record all behaviors that are relevant, researcher writes notes

Overt and covert Participant and non-participant

Structured and unstructured Naturalistic and controlled

Overt : participants know

Covert : participants do not know

Participant observation : researcher joins group

Non -participant observation : researcher observes from a distance

# 01 02

# 03 04

# Observation (2/5) Evaluating observation

# Observation (3/5)

Advantages Disadvantages

Structured observations are more reliable than unstructured observations as the behaviours that are to be recorded are decided in advance and operationalised. Behaviour that is being observed will be broken down in components that can be easily identified and recorded. This ensures that different observers are more likely to record data consistently .

Structured observations can also be subject to observer bias as they rely on the observer's own judgement as to what behaviour to record and are therefore subjective.

Unstructured observations can be used as part of a pilot study to give researchers a good understanding and overview of the range of behaviours that they might observe. The researchers would then choose a limited number of these to observe for their main research.

In unstructured observations, it can be difficult to record all behaviour

and some of the subtler (but interesting) behaviours might be missed. Evaluating observation

Advantages Disadvantages

An advantage of overt observations is that it is

more ethical because participants know they are being observed so it avoids having to deceive participants.

A disadvantage of overt observations is that participants may show demand characteristics

because they know they are being watched and therefore may not act in a natural way

An advantage of covert observations is that it

avoids any potential demand characteristics

because the participants do not know they are being observed.

A disadvantage of covert observations is that they

are less ethical than overt observations because participants do not know they are being observed and studied. Evaluating observation

Advantages Disadvantages

An advantage of non-participant observations is that the observer is likely to be more objective in their observations as they are not personally getting involved in the study.

A disadvantage of non-participant observations is that because the observer is watching from a distance some behaviours may be missed.

An advantage of participant observations is that the observer is not viewing from a distance and therefore may gain a greater understanding of the participants' behaviour.

A disadvantage of participant observations is that the observer may become too involved with the people that they are observing, and become less objective in their observations Evaluating observation

Advantages Disadvantages

An advantage of naturalistic observations is that behaviour is likely to be normal as the participants are in their own natural settings and the researchers do not interfere in any way. Therefore it is likely to be high in ecological validity.

A disadvantage of naturalistic observations is that it is much harder to control for variables that might affect the participants' behaviour.

An advantage of controlled observations is that they can be more easily replicated by other researchers as they can use the same behaviour schedule and it is easier to standardise the situation for all participants. This makes controlled observations more reliable.

A disadvantage of controlled observations is that behaviour may be less natural if participants are aware that they are in a controlled setting. Researchers might also miss key behaviours if they have a very rigid behaviour schedule.  Looks at the relationship between 2 variables

These are called co-variables

For each participant, researchers will gather 2 sets of data which they can plot on a scatter graph to see whether there is a correlation between 2 measures/variables

# Correlations (1/8) Operational definition:

Operationalizing the variables means to state exactly what they are and how they will be measured

Example: Co-variable 1: average time spent on playing video games per week  measures in hours Co-variable 2: reaction speed on an app

measured in minutes, seconds, and milliseconds It is ALWAYS important to operationalize hypothesis for correlations

# Correlations (2/8) Operational definition:

Correlations are often used by researchers:

To investigate new areas of psychology before conducting experimental research,

OR in cases where it is not practical or ethical to manipulate variables

Various methods used for data collection, AS LONG AS THE DATA IS LONGITUDINAL

Data is analyzed by plotting pairs of scores in one another on a scatter graph (to see whether there is any correlation or not)

# Correlations (3/8) Positive, negative, and no correlations

Positive correlation  2 variables INCREASE together

EXAMPLE : Dement and Kleitman (sleep and dreams) - the longer the duration of REM sleep (in minutes) the more words participants used to describe their dream.

# Correlations (4/8)  Negative correlation  When one variable INCREASES, the other variable DECREASES

EXAMPLE : In Baron-Cohen et al. (eyes test), the higher the Eyes Test score the lower the Autism Spectrum Quotient score

# Correlations (5/8)  No correlation  no consistent relationship between scores

EXAMPLE : No correlation between time spent on mindfulness activities between the weekly MBSR sessions and grey matter concentration.

# Correlations (6/8)  Strength of the relationship between 2 variables can be represented by a CORRELATION COEFFICIENT .

In general, the closer to 1 (for a positive correlation) OR the closer to -1 (for a negative correlation), the STRONGER the correlation is.

IS CORRELATION COEFFICIENT VALUE OF +0.75/-0.75 A STRONG COEFFICIENT?

# Correlations (7/8) Evaluating correlations

# Correlations (8/8)

Advantages Disadvantages

Correlations allow psychologists to investigate new areas of research to see whether they are worth investigating further with experimental methods .

Correlations can only show the strength of the relationship between two variables and cause and effect cannot be inferred. They also do not tell us why that relationship has occurred

Correlations allow researchers to investigate areas where it is not practical or ethical to manipulate variables.

The presence of a third variable cannot be ruled out in correlations. A third variable that is not being measured may have influenced the relationship . Importance of objective data collection  Involves following the same group of individuals over an extended period of time and giving them tests or tasks at various intervals during that time to see how they change and develop in their thoughts, feelings and/or behaviour.

Participants of longitudinal designs need to be re-contacted over time for repeated testing.

This testing needs to be standardised for valid and reliable comparisons to be made between participants' scores over time.

# Longitudinal studies (1/3)  Longitudinal studies use a variety of techniques to gather data , from psychometric testing to interviews and questionnaires.

Experiments can also use a longitudinal design where the experimental group is given an intervention (such as a wellbeing programme, an educational intervention or a dietary supplement) and compared over time to a control group that is not given the intervention.

Follow-up sessions allow psychologists to measure the long-term impact of treatments and interventions on the participants.

# Longitudinal studies (2/3) Longitudinal studies (3/3)

Evaluating correlations

Advantages Disadvantages

Because longitudinal studies are following the same group of participants as they change and develop over time, researchers do not have the issue of

participant variables and participant effects

Due to the long-term nature of the research, it is more prone to attrition  where participants drop out of the research, either because they move away and lose contact with the researchers, they decide they no longer wish to take part or their life circumstances change.

Allow some topics to be studied which might not be possible using other research methods, such as child development over time.

By the time longitudinal research is completed, it may no longer be generalisable due to changes in society This means it may lack temporal validity Group work

Make your own experiment for one of the following topics

1. Social anxiety and baron-cohens eye test for social sensitivity

2. Differences in intelligence between millennials genZ

3. Bullying and depression

4. Childhood trauma and adult depression

Your explanation should include

1. Type of research design (correlational/ experimental/ case study ect)

2. Type of data (quantitative/qualitative)

3. Procedure

4. Type of instruments used (interview/ questionnaire)

5. The results you expect

6. Weaknesses and strengths (ethics included) Aims

The aim is a statement that explains why the researcher is doing the research and what they hope to achieve. Hypotheses

A hypothesis is a statement where the researcher predicts what they think will happen in the research. Hypothesis is operationalised; meaning that the independent and dependent variables (in an experiment) or the co-variables (in a correlation) are clearly defined. The units of measurement (e.g. milliseconds) should always be included. For example, a phobia could be measured using a score from a self-report questionnaire, heart rate from a heart rate monitor (in beats per minute), a participant's body language from an observation (e.g. number of times they touch their face or hair, cross and uncross their legs, etc.) or a measurement of cortisol (a stress hormone) in their saliva. Directional hypotheses

With a directional hypothesis:

The researcher will predict whether the experimental group will perform significantly better or worse than the control group (or in a correlation, whether it will be a positive or a negative correlation).

Directional hypotheses tend to be used where previous research has been done in the area, so the researcher has a good idea of what results to expect Example: People with childhood trauma will develop depression in adulthood People who drink coffee will have quicker reaction times than people who drink a cup of water Non-directional hypotheses

For a non-directional hypothesis:

Researcher states that there will be a significant difference between the experimental and control groups but does not state in which direction that difference will be .For example, whether the experimental group will be faster or slower. Similarly, with a correlation, the researcher will state that there will be a significant correlation between two variables, but not whether it will be a positive or negative correlation . Non-directional hypotheses

Researchers tend to use non-directional hypotheses:

When there has not been previous research conducted in the area that they are researching

When the research is contradictory regarding what the possible outcome might be Null hypothesis

The null hypothesis is the statement that the independent variable will have no effect on the dependent variable.

That two variables are not correlated with each other. It is important for the null hypothesis to also be operationalised. Null hypotheses normally start with 'There will be no difference/correlation ... ', and they usually finish by saying that 'any difference that does arise will be due to chance' Data analysis

Psychologists analyse quantitative data using descriptive statistics These help to summarise the data and make it easier to see whether there are any obvious patterns or trends. Psychologists use a branch of mathematics called inferential statistics, which allows them to calculate the probability (chance ) that their result would have arisen if the null hypothesis was true. This is called a p value. Having completed their analysis, the researchers will decide whether to accept or reject their null hypothesis. The null hypothesis is rejected if the p value (the probability of a significant result having occurred due to chance) is less than 0.05 (1 in 20) and it is accepted if it is more than 0.05 (1 in 20). Controlling of variables

When conducting experiments, researchers attempt to control all variables other than the one being manipulated (the independent variable)

Controlled variables are 'held constant' , or standardised , between the two or more conditions

This allows the researcher to state that any changes in the dependent variable were caused by the independent variable . Standardisation of a procedure

In order to make research replicable so that it can be checked for reliability, it is essential that the procedure is standardised for all participants.

In an experiment, the only difference should be what the researcher is manipulating (the independent variable).

However, all research, whether it is experimental or not, should follow standardised procedures.

This includes participants all receiving the same information and instructions (unless this is something that is being manipulated as part of the independent variable) . Participant variable

Participant variables are aspects of a person's background , personality ,

cognitive abilities , health and wellbeing , etc., that affect how they behave or respond in a study.

Individual differences between people are of course inevitable and this is why

psychologists work with averages calculated from larger groups of people.

However, participant variables can cause problems in experimental designs if some aspect of the participants (e.g. empathy, aggression or intelligence) varies systematically between the two or more experimental and control groups. Participant variable

For example; if all passengers in Piliavin et al.'s 'cane' condition were high in empathy and all the people in the drunk condition were low in empathy, it would be impossible to know whether aspects relating to the victim (drunk or cane) affected helping behaviour, or whether it was individual differences (high and low empathy) between the passengers that affected behaviour. How to overcome participant variables?

1. Random allocation of participants to the experimental and control groups can help overcome the effects of participant variables on the results. This was not possible in Piliavin et al.'s study as it was a field experiment

But

In Andrade's study (doodling) participants were randomly assigned to the experimental (doodling) and control (non-doodling) groups to ensure that people with naturally good memories did not all end up in the doodling group How to overcome participant variables?

2. Using a matched-pairs design can help reduce the impact of participant variables, as researchers can match participants in both the experimental and the control groups on key factors that could impact on the research (age, handedness, cultural background or socioeconomic status) For example, in Bandura et al. (aggression) participants were matched on their existing aggression levels, so that each group (experimental and control groups) contained children with lower and higher levels of aggression. Again, if the most aggressive children had all ended up in one of the experimental groups, it would be unclear whether they bashed Bobo due to observing an aggressive adult or they would have done this on their own anyway. Situational variables

Situational variables refer to any aspect of the environment that could impact participants' behaviour and affect the results.

Environmental variables include factors like:

Temperature of the room

Lighting and noise levels It is important for researchers to control these variables as much as possible so that each participant experiences the same environment, AND it is ONLY

the independent variable that is affecting the dependent variable. It also makes the research replicable Uncontrolled variables

These are factors that are not being measured or controlled by the researcher and can have an unwanted effect on the dependent variable. These could be things such as the temperature of the room or noise levels. Types of data

Quantitative data refers to any data that is

numerical . For example, closed questions on questionnaires produce quantitative data

Qualitative data refers to non-numerical data, usually in words but sometimes in images (e.g. photographs). Open questions produce qualitative data as participants are able to freely express their thoughts and feelings. Types of data

Strengths Weaknesses

Qualitative data

Qualitative data provides researchers with a more in-depth understanding of what their participants are thinking and feeling and the reasons behind their behaviour, which may give the data greater validity

Qualitative data analysis may be more at risk of researcher bias as the interpretation of the data is more subjective . Therefore, the findings can be less reliable.

Quantitative data

Quantitative data can be more easily compared.

The analysis of quantitative data is

objective and so less prone to researcher bias.

Quantitative data does not always allow us to fully understand what a participant is thinking or feeling . Some scales may limit how a participant can respond and they may feel that none of the responses reflects how they feel, making the data less valid. Types of data

Subjective and objective data

Subjective data means that the data can be influenced by a person's personal thoughts, feelings or opinions. This is more likely to be an issue with qualitative data, which requires interpretation by the researcher when being analysed.

Objective data is unbiased, factual and not influenced by a person's personal thoughts or opinions. For example, research that uses ratings scales that are quantifiable is more objective as the data does not require interpretation by the researcher. In scientific research, it is important to be as objective as possible. Researchers can improve objectivity by getting another researcher, who does not know the aim of the study, to interpret the data. This person is likely to be more objective than the primary researcher, who will have already formed hypotheses about what the data will show. Types of data

How to improve objectivity?

Researchers can improve objectivity by getting another researcher , who does not know the aim of the study, to interpret the data. This person is likely to be more objective than the primary researcher, who will have already formed hypotheses about what the data will show. Sample and population

Sampling refers to how researchers obtain participants for their research. Researchers will have a target population of people to whom they wish to generalise their findings. For example, the target population might be the global community of Cambridge International A Level psychology students. It would be practically impossible to test the entire population and instead the researcher selects a sample from the target population. Collectively, researchers refer to their participants as the sample .Sampling of participants

Sampling techniques

1. Opportunity sampling: There are a number of ways that researchers can select their sample. Opportunity sampling is a technique in which researchers select participants who are readily available. For example, a psychologist might know six teachers from different countries around the world who teach Cambridge International A Level. The psychologist will ask those teachers if they can use their students. Many research studies are conducted by university lecturers using their own students. Sampling of participants

2. Random sampling: With random sampling, every person in the target population has an equal chance of being chosen to participate in the research. For example, if the target population is Cambridge International A Level students, this means that each Cambridge student across the world would have an equal chance of being selected as a member of the sample. This would involve getting the names of all of the students and putting them into a random name generator (computer program), which would pick a sample of 100 students, for example. Sampling of participants

3. Volunteer (self-selecting) sampling: participants put themselves forward to take part in a piece of research. Researchers may advertise for participants in many ways including posters, flyers, direct mail, e-shots, newspapers, radio and online forums. For example, if the researchers were looking to study a condition such as prosopagnosia (a condition where people struggle to recognise faces), they could find a Facebook group of people with the condition and advertise their research in the group so that members can volunteer (self-select) themselves to take part. Sampling participants

Strengths Weaknesses

Opportunity sampling

Quick and easy as researchers use people who are readily available at the time of the research. May result in a larger sample and likely to mean research can be conducted without delays.

Often unrepresentative of the target population . Findings may not be generalisable.

Volunteer sampling

Participants have volunteered to take part so they are motivated and willing -drop-out rates are likely to be low . This is important in longitudinal research where attrition may significantly affect validity

May not be representative as only certain personality types volunteer to take part in research , such as those with more pro-social attitudes, those who are more motivated than the rest of the population and those who have a personal interest in the research topic. Sampling participants

Strengths Weaknesses

Random sampling

Should be representative of the target population , making the findings more generalisable (if the sample is large enough).

Time-consuming and expensive .

Potential participants may not wish to take part in the study . All potential participants must be contacted and arrangements must be made for them to visit the location where the research is being he Ethics and human participants Ethics

Ethics are critical component of all psychological research Wherever in the world there is psychology, there are always ethical guidelines Informed consent and right to withdraw- core principles of ethics Ethical guidelines:

Serve to protect participants

Ensure they do not come to any hard

Ensure that research that benefit wider society can be pursued Valid consent:

This means participants know exactly what will be happening during the research before they agree to take part

Consent must be provided before the data collection begins BUT is not considered valid until it is INFORMED

As in, participants must know what they have signed up for and should be given a detailed, age-appropriate information Researchers must be able to provide evidence that consent was obtained Valid consent:

Incase of some vulnerable groups, parents family members, or caregivers should also give consent

Eg: consent from parents or caregivers when working with adults with low cognitive and communication abilities Parental consent must be obtained when working with children (under 16 or under 18 depending from country to countrys age restrictions) RIght to withdraw:

Although participants give their consent, they must also be aware that they can withdraw from the research at any time without any pressure

They must be informed that their data will be deleted or destroyed

Researchers working with very young children or adults with significantly impaired communication ability should look for signals (verbal and non-verbal) that participants is no longer to participate Minimizing harm and maximizing benefit:

Must aim to minimise harm to participants psychological well-being, personal values, privacy or dignity and mental health

Risk of harm should be no greater than the participant might expect to experience in their everyday lives

Should aim to maximise the benefits of their research at all stages of the research process Lack of deception:

By obtaining informed consent, participants will be very clear about what they will be doing and the aim and purpose of the study

Lack of deception means that the participants will not be deceived about anything about the research

In reality, it will be difficult to conduct research without any deception as simply knowing the true aim of the research could change participants behavior

When deception is involved, full debriefing is essential confidentiality:

Participants results should remain confidential before and after publication of their data

Particularly important when the data obtained is qualitative; more likely to be personal and thus more identifiable

Data should be kept confidential while conducting the research and analysing the data

Participants names should not be included with their data Privacy:

In the context of observational research, participants should only be observed in public situations where they might be expect to be observed by others

Not acceptable to observe people in their own homes without consent Debriefing:

Happens after the participants has completed the research

Purpose of debriefing:

to explain any deception that has taken place

to answer any questions the participants may have

To ensure they leave the research in the same state in which they entered in (negative state of mind t should be turned to positive state of mind) Ethics and non-human participants Minimizing harm and maximizing benefit:

Must seek to minimize harm, discomfort, and suffering to the animals

Maximize the benefit of the research in applying their findings to helping other animals or animals of the same species

Replacement:

The use of animals is replaced with video footage from prior research or computer simulations

Particularly important if animals are to be used for training or teaching purposes Species:

Must important to use species or animals that are both scientifically and ethically suitable for the research

Researcher:

must have good understanding of the animals natural history and its ability to feel

Researchers should know whether animals have been bred in captivity or not

Researchers must choose the species that is likely to suffer the least Numbers:

Must use the smallest number of animals possible to meet the research aim

Use of pilot studies with small number to ensure that design is appropriate and testing what the researchers want to investigate before using a large sample of animals

Good experimental design ensures the minimum use of animals Procedures- Pain and Distress:

Should avoid causing death, diseases, psychological or physical discomfort to animals

Should look at enriching animals environment rather than depriving

Should consider balance between harm caused to the animals and the benefits that would be gained

Pain is ONLY ETHICALLY JUSTIFIABLE:

when there is no alternative research method available

research would have significant scientific or educational value. Surgical procedures should be conducted using anaesthesia to reduce pain and prevent infection If termination is needed, must be done swiftly Housing:

Social and natural behavior of the species should be considered when housing the animals

Isolating animals who would normally live in social groups could cause stress

Overcrowding could also cause stress and aggression

Using same cages could reduce negative impact Reward, deprivation, and aversive stimuli:

If researcher is considering use of food deprivation, they must have a good understanding of the species normal eating and drinking habit to ENSURE THAT DEPRIVATION IS KEPT TO A MINIMUM. Validity Validity

Extent to which the researcher is measuring

what they think they are measuring and the

extent to which the findings are useful and meaningful Historical example of a psychometric test that lacked validity

> Army Alpha and Beta IQ tests:

The term 'IQ' stands for ' intelligence quotient'. In 1917, a psychologist called Yerkes worked with colleagues to create IQ tests for army recruits in the United States of America, to see which section of the army would be most appropriate for them, based on their IQ levels. Many of the recruits were immigrants who had not lived in the country very long. However, the 'IQ' tests contained questions such as 'What is Crisco?' and 'Who is Christy Mathewson?'. These questions were impossible to answer unless the recruit had a good knowledge of American culture. This led to recruits from certain nations experiencing prejudice as they (understandably) received low IQ scores. This is a good example of a test that lacks validity - it was testing cultural knowledge, not IQ . Ecological validity

The extent to which behaviour that participants demonstrate during research relates to how they would behave in their real lives. Research conducted in normal, real-life settings , as in field experiments, is likely to be

higher in ecological validity .However, some laboratory experiments can also be high in ecological validity if the setting is unlikely to impact on behaviour. Ecological validity

> If research lacks ecological validity , then it cannot be generalised to real-life situations. However, where it is important for research to be conducted under controlled conditions, elements of the procedure can be made more ecologically valid.

For example, if participants had to demonstrate their driving skills in a laboratory setting, it would have greater ecological validity if a participant were able to use virtual reality headsets with a steering wheel, brake and accelerator rather than using a computer keypad as controls. Subjectivity/Objectivity

If a researcher is required to interpret behaviour, or the data collected, in any way, there is always a risk of subjectivity . This is where their personal thoughts ,

feelings or opinions may affect the validity of the findings.

An example: researcher misinterprets a certain behaviour and records it in a way that was not meant. This would lead to the data not being valid. In terms of data analysis, qualitative data is most at risk of subjectivity as it requires a researcher to analyse the meaning behind the data and draw conclusions based on this.

# Evaluating studies based on their validity Subjectivity/Objectivity

> The researcher's own interpretation may be influenced by their personal biases , either consciously or
> unconsciously , leading to low validity.

For example, Milgram (obedience) needed to interpret the behaviours of the participants as they delivered what they believed to be electric shocks to another person and decide what data to record. Having previously conducted a pilot study, he may have already had certain expectations about what behaviours to expect. Subjectivity/Objectivity

> If data collection or analysis is not influenced by a researcher's personal thoughts, feelings or opinions , then it is said to be objective .

Scientific equipment - for example, brain scanning techniques such as EEGs used in Dement and Kleitman (sleep and dreams) and MRis used in Holzel et al. (mindfulness and brain scans) - is only capable of recording what biologically occurs, thus this method of collecting data is completely objective. Subjectivity/Objectivity

It is worth noting , however, that the images produced by brain scans still require the subjective interpretation of researchers . Quantitative data is also more objective than qualitative as it can be statistically analysed. The less risk the data has of being influenced by another person's interpretation, the more objective and some would argue, the more valid. Demand characteristics

Participants might try to guess the aim of the research and change their behaviour based on what they think the researcher is looking for. This would reduce the validity of the study. Any features of the research environment which may give away the study's aim are called demand characteristics. Demand characteristics

This is why some studies use deception to avoid the participants' behaviour being affected. In Milgram (obedience), participants were deceived in a variety of ways to attempt to hide the study's true purpose. One of the strengths of using animals as participants is that they do not actively try to work out what the study is about . Therefore, demand characteristics are reduced and validity increased. This said, animals may unconsciously pick up on subtle cues from researchers, which may affect their behaviour. Generalisability

Findings should only be generalised to the target population IF the sample was sufficiently representative.

This will depend upon the sampling technique, for example random samples are typically more representative than opportunity or volunteer samples. The extent to which findings can be generalised beyond the sample is sometimes called 'population validity' .For example, the study by Piliavin et al. (subway Samaritans) looked at helping behaviour on a New York subway carriage. It may not be appropriate to generalise their findings to all cultures because helping behaviour may differ in other countries, based on cultural values. Generalisability

> Whether the results from a study can be generalised to real-life situations or not is called ecological validity.

For example, as the study by Piliavin et al. (subway Samaritans) took place on a normal subway journey, as part of people's real lives, it is possible to generalise the findings to similar real-life contexts Temporal validity

> Refers to whether the results of a study can be generalised to a different time period.

For example, Bandura's study was conducted in the 1960s and the results may not be the same today as modern children have access to personal technology such as phone and tablets and gender roles have also changed significantly. This means the findings may be era-bound and said to lack temporal validity . Other studies, such as Hassett's study of sex differences and play in monkeys, are less likely to be affected by temporal validity since they are based on a generally stable biological functions. Social psychological research may be most susceptible to issues with temporal validity, as cultural and societal norms change over time . Reliability

Reliability is about consistency and can refer to the results from each participant and the overall findings, based on the whole sample. If you tested your participants again and the results were the same, we could assume the results were reliable. If we ran the whole study again, with a different set of participants (a replication) , we could check the reliability of the findings. If we got the same results again , we could conclude that these findings were reliable.

# Reliability and replicability Replicability

Replication means that other psychologists can repeat the study exactly to see if they get consistent results.

Therefore, the ability to replicate a study is critical as it helps researchers to demonstrate the reliability of their findings - that is, to show that their findings were not due to chance. Replicability relies heavily on the extent to which the study has a standardised procedure that other researchers can follow in their replications.

# Reliability and replicability Standardization:

Standardisation of the procedure and instructions provided to the participants is an important way of making a study replicable. It also means that all aspects of the research are consistent for every participant Standardization

For example, in the study by Bandura et al. (aggression), the model always displayed the same sequence of behaviours, in the same order and for the same length of time. The toys were laid out in exactly the same way in each of the rooms and the children were always given the exact same standardised instructions. The observer also scored the children's aggressive and non-aggressive behaviour in the same way according to the categories on the behavioural checklist. This high degree of standardisation means Bandura's study is easy to replicate and, therefore, his findings could be checked for reliability. Ways to increase reliability Owari 1. Describe the ethical guidelines of 'species and 'replacement' in relation to animals in research, using any examples (6m) 2. Mirabel is observing two groups of gorillas in the wild, a large group of 65 and a small group of 20. She is recording the number of close social relationships each gorilla has with other gorillas in the group. Outline why Mirabel must have a strong knowledge of the 'species before being allowed to conduct this work (2m) 3. lchika is conducting a study on recall of items from a menu. A waiter reads out a list of nine dishes to diners in a restaurant. They are then asked which dish they would like to order. lchika records whether they choose from the first, middle or last three menu items. She allows half of the participants to doodle on a napkin while they are listening to the waiter and the other half are not allowed to doodle. Explain one reason that lchika's study has high validity and one reason why it has Low validity (4m) 4. Explain what is meant by the terms 'inter-rater reliability and 'inter-observer reliability using any examples (6)

Transcript for:Overview of Research Methodology Essentials

Transcript for:
Overview of Research Methodology Essentials