Research designs fall into 3 broad categories



Introduction to Psychology

Revision Summary

[pic]

Key Ideas

• Empirical investigations in psychology may be experimental, quantitative observational, or qualitative.

• All investigation designs and methods of assessing psychological responses have advantages and disadvantages.

• All research involving humans has ethical dimensions.

• Different types of representation are appropriate for different types of data.

Areas of Learning

• The range of investigation designs that can be used to answer a particular research question and their advantages and disadvantages; the three investigation designs used in psychology — experimental, quantitative observational, and qualitative; focus groups and the Delphi technique as examples of qualitative investigations; advantages and disadvantages of quantitative and qualitative investigations; the difference between experimental investigations and quantitative observational investigations in their design.

• The three methods of assessing psychological responses — objective quantitative measures

(e.g. physiological measures such as heart rate, behavioural counts, scores on standardised intelligence tests), subjective quantitative measures (e.g. responses on checklists and rating scales, scores on personality tests), and qualitative assessment of data; content analysis of responses in focus groups as an example of qualitative assessment; awareness of the limitations of drawing conclusions using small or unrepresentative samples; consideration of the validity and reliability of the methods.

• Descriptive statistics (i.e. the ways in which quantitative and qualitative data may be represented and described); the generalisation of research findings. Instruction in statistics should be limited to determining medians and means; generating graphical representations of data; interpreting medians, means, standard deviations, and graphical representations of data. However, a brief description of the function of inferential statistics and criteria for significance will enable students to read original research with some understanding.

• Ethical issues associated with investigations; the ethical safeguards that have been incorporated in particular investigations.

In addition to these points you must also understand:

This subject is designed around the four levels of explanation of behaviour used in psychology. Behaviour can be explained in terms of biological processes, basic psychological processes, the attributes of the person enacting the behaviour, and sociocultural processes. These four levels are explained as follows:

• The biological level of explanation focuses on the biological and chemical processes underlying behaviour. For example, a biological level of explanation for interpersonal aggression might focus on the role played by hormones or activity in specific brain areas.

• The basic processes level of explanation focuses on the psychological processes that are universal (or at least very widespread) across humans. For example, a basic processes level of explanation for interpersonal aggression might focus on the cognitions and emotions that commonly precede it.

• The person level of explanation focuses on individual differences in behaviour. For example, a person level explanation for interpersonal aggression might focus on different levels of aggression displayed by persons with different types of personalities.

• The sociocultural level of explanation focuses on the influence that other people exert on behaviour by studying behaviour in social and cultural contexts. For example, a sociocultural level of explanation for interpersonal aggression might focus on the role played by onlookers or on different levels or types of interpersonal aggression displayed in different cultures.

Research designs fall into 3 broad categories

All allow empirical data to be collected

• Experimental: researcher manipulates variable of interest – e.g., assigns participant to

• Consume caffeine Vs no caffeine

• recall real Vs nonsense words

• Quantitative Observational: capitalise on pre-existing differences in variable of interest

• Younger Vs older adolescents

• Low Vs high resting heart rate

• Qualitative: --> rich verbal data, in own words

– e.g., focus groups, Delphi technique, unstructured or semi-structured interviews.

Experimental designs

– Essential feature = Researcher manipulates some factor thought to influence behaviour

• “independent variable”

– Can investigate

– Differences between groups

– Associations between variables

– Single cases

– One of the most controlled research methods used in psychology.

– Used to test a cause – effect relationship.

– Variable – any factor that can vary in quantity or quality.

o Independent variable – is systematically manipulated or varied in some way by the experimenter.

o Dependent variable – shows whether or not the manipulation of the IV has had any effect

o Extraneous variable – is a variable that can cause changes to occur in the DV other than the IV

▪ Random in nature rather than systematic, environmental, participant differences

Features of the experimental design

The experimental method usually uses two different groups called experimental groups and control groups. In the simplest experiment there would be one experimental group and one control group.

Experimental group

Participants in the experimental group are exposed to the independent variable (that is the group that receives the “treatment”. The treatment (IV) is the hypothesised cause of any variation in the dependent variable (the hypothesized effect.)

Control group

The control group consists of participants who are not exposed to the treatment (that is the independent variable is absent) They provide baseline data against which the experimental group data can be compared. This comparison helps to determine whether the independent variable has caused some change in the behaviour (the dependent variable)

It is extremely important that the experimental group and the control group are as similar as possible in relevant participant characteristics (representativeness) and that they are treated as similarly as possible throughout the experiment.

Pre-testing and post-testing

Both groups are tested at the beginning of the study and at the end. The results of the experimental group are then compared to those of the control group to see if the treatment made a difference

Random allocation

If allocation is by random assignment, experiments allow conclusions about cause

Random allocation means that all participants who have been selected for an experimental are just as likely to be in one group as the other. This random allocation may be performed by coin flipping or some other procedure that leaves the outcome entirely up to chance.

When the sample size is large enough, it can be expected that each group will contain roughly equal numbers of relevant participant characteristics. The purpose of random allocation is to form groups that are as equal as possible on all characteristics except the independent variable.

Advantages and Disadvantages of the Experimental Design

Advantages

– As well as controlling the IV the experimenter attempts to eliminate unwanted extraneous variables

– Control over extraneous variables is usually greater than in other research methods

– Experimental design involves manipulating the IV to observe the effect on the DV. This makes it possible to determine a cause and effect relationship.

– Because of strict conditions and control the experimenter can set up the experiment again and repeat or ‘check’ their results. Replication is very important as when similar results are obtained this gives greater confidence in the results

Disadvantages

– Elimination of extraneous variables is not always possible

– Experimental situation may not relate to the real world

– It may be unethical or impossible to randomly assign people to groups.

o Eg unethical to deliberately deprive children of sleep

o Impossible (expensive) to observe jetlag by sending people on world trips

o Unethical to experiment on forms of punishment on children by subjecting them to punishment

Quantitative Observational Designs

Essential feature = Researcher observes outcome of natural variation in factor of interest

• “predictor” variable

• Can investigate

– Differences between groups

– Associations between variables

– Single cases

Quantitative Observational Designs have many features of a true experimental design, but don’t have the same level of control over extraneous variables.

Differences between Experimental Designs and Quantitative Observational Designs

One of the main differences is that the ‘treatment’ pre-exists in the two groups being compared whereas in an experiment the researcher manipulates the research groups.

– Eg, finding out whether gender affects memory (gender is pre-existing – born male or born female)

– Finding out attitudes to smoking (randomly allocating to one of two groups)

Advantages and Disadvantages of the Quantitative Observational Design

Advantages

– Quantitative Observational designs allow variables to be investigated that would be unethical, impossible or too costly under an experimental design.

– Some kinds of behaviour can only be observed in a naturalistic setting (eg unethical to deprive children in early life to observe effects later in life)

Disadvantages

– Cannot infer such a strong cause and effect relationship because there is a greater chance of other variables affecting the results. This is due to the lack of random assignment to groups.

– Cannot replicate the findings as the same situation will not occur naturally again

– Q.O. does not allow generalisation of findings to the general population

– Observer bias may influence the results

Qualitative

Essential features = aim to generate rich verbal responses and aim to limit the constraints on the nature of the data (ie give people the chance to respond as they desire, rather than limiting their responses.

A qualitative research method enables more valuable understandings of behaviour, as experienced in real life, compared to precise numerical data achieved through quantitative research.

The focus of qualitative research is to find the answer to a general research question (What are the effects of X?), rather than testing a specific hypothesis (X leads to an increase in Y?)

Can investigate

– Differences between groups

– Associations between variables

• Some qualitative designs (e.g., Delphi technique, focus groups) can generate both quantitative and qualitative data

Focus Groups

➢ Focus groups refer to a group interview technique that obtains data through discussion between research participants in a group setting.

➢ Participants are encouraged to talk to one another, ask questions, exchange personal experiences and comment on each others experiences and opinions.

➢ Make use of focussed, open ended discussion lead by a trained facilitator

Advantages and Disadvantages of Focus Groups

Advantages

– May be more convenient that one-on-one interviews

– Rich data can be collected

– Easy to organise and inexpensive to conduct.

– Useful for collecting information from people who are unable to read or write

– Can encourage participants who would normally not speak out or have a opinion or reveal information about themselves

Disadvantages

– People may not feel comfortable expressing opinions in front of others

– A few people may dominate the discussion

– Interviewer bias may contaminate the results

– Cannot establish causation

– The data from a focus group should not be interpreted to be representative of the entire population

Delphi Technique

➢ The Delphi technique uses a series of self-administered questionnaires and feedback to obtain the opinion of experts in a field of interest.

➢ Generally involves two steps.

▪ Relevant individuals are invited to provide open-ended opinions on a specific topic

▪ The opinions from step 1 are grouped together into themes in a questionnaire then redistributed amongst participants as a self-administered test. Participants are usually required to rank their agreement with each statement in the questionnaire.

▪ The process is repeated until a consensus is reached.

Advantages and Disadvantages of Delphi Technique

Advantages

– Participants are free to express their opinions due to confidentiality and anonymity of their responses.

– A greater pool of expertise may be tapped into compared t individual interviews

– A small or large group of experts can be contacted cheaply through post or email, not limited by Geographic’s (ie can send questionnaire all over the world)

Disadvantages

– Researcher bias can contaminate results

– Does not allow for generalisation of results due to the limitations of using a small sample group

– May force consensus as people are not able to discuss responses (researcher surmises what the people felt overall)

Comparing Quantitative and Qualitative investigations

Qualitative research can provide a richer and deeper understanding of participants and their responses, however factors such as researcher biases can influence the results. Social desirable responses may also be given.

Quantitative research usually involves data collection under strict conditions to prevent or minimise the influence of such factors.

If research is about meanings and experiences, then qualitative research will provide more meaningful data

If research has a more restricted focus and is about people’s behaviour quantitative research will be more meaningful

Three methods of assessing psychological responses

Objective Quantitative (numbers)

– Recording physiological responses

• EEG (Brain activity during sleep), heart rate, functional MRI

– Measurement of behaviour

• e.g., response latency, frequency of specific behaviour, number of errors

Subjective Quantitative (numbers)

• Self-reports – e.g., attitudes, self-reported behaviour

• Reports by others

Qualitative only (words)

– Content analysis

Objective quantitative measures

Objective = verifiable

Quantitative = numerical

Physiological Responses

Examples:

• Brain activity

– Electroencephalograph (EEG)

– Functional Magnetic Resonance Imaging (fMRI)

• Psychological arousal

– Heart rate

– Galvanic skin response

– Blood cortisol levels

Behavioural counts

Using behavioural counts is an observational method for which a researcher views a group of people or animals to record the number of times a particular behaviour occurs

Eg observing kindergarten children through a one way mirror

Scores on standardised tests

Scores on standardised tests are used to measure the individual differences that exist among people. Common examples are intelligence tests. This also allows researchers to compare responses against a normal distribution graph. An individuals score on a standardised test allows the researcher to determine what proportion of people can be expected to score above or below that individual.

They are standardised as they are administered and scored the same way each time.

Subjective quantitative measures

(e.g. responses on checklists and rating scales, scores on personality tests)

Subjective = guess or estimate

Quantitative – numbers

These measures also gather numerical data, but the researcher cannot be as confident that participants have provided accurate data free from bias. Participants may feel inclined to give socially desirable response.

Examples:

Rating scales – typically a series of fixed-response questions about different aspects of a research topic (eg attitudes to racism, feelings about sleepiness) which require participants to “rate” on a scale their thoughts at that time.

– Can be a Likert scale – strongly agree, agree, neither, disagree or strongly disagree.

– Can be a rating scale from 1 – 10

– Can be placing a mark on a long line to indicate % of support ___________________ (=70%)

Questionnaires – with fixed response answers (eg yes/no answers, multiple choice answers or rating scales) They allow the opportunity to ask the same questions of a large portion of people.

Qualitative measures

The Delphi technique and Focus Groups are methods for collecting qualitative data.

These data are subjective as the researcher infers meaning as they interpret the data. Allows for individual interpretation of the data. Two individual researchers may come up with different summaries or interpretations

Examples:

Content analysis of responses in focus groups

Researchers follow the following steps to organise the data gathered from focus groups.

1. organise the data become familiar with the data, read and re-read

2. identify core themes (identify groups of comments that are similar

3. code themes – develop a name that identify’s each theme)

4. keep track of themes – note themes that are recurring and whether they were the first comment made or in response to another comment look at agreements, disagreements, contradictions, different ways of saying the same things

5. analysis – summarise the analysis using a frequency table with illustrative quotes

|Themes |Frequency of themes |Examples from participants |

|Emotions when tired |6 |Grumpy |

| |4 |Angry |

| |3 |moody |

|Behaviours |3 |Lack of concentration |

| |4 |Sleep |

|Times when tired |3 |Morning |

| |2 |Lunch time |

| |2 |After lunch |

| |1 |5 pm |

Limitations of drawing conclusions using small or unrepresentative samples

A sample is usually a subsection, or smaller group, of research participants selected from a larger group (population) of research interest.

The term population refers to the entire group of research interest from which a sample is drawn.

Whether carrying out a survey or an experiment it is important that the sample is a random sample of the population. In other words every person in the population had an equal chance of being chosen in the sample. If it is not a random sample the experiment or survey may not accurately reflect the population.

Size: larger samples are more likely to reflect the population

Unrepresentative samples: an unrepresentative sample is a sample which does to represent the population due to its size or that is biased in term of factors such as gender, age, socioeconomic status or cultural groups. An unrepresentative sample can lack internal validity which means that it does not test the hypothesis.

Using small or unrepresentative sample doesn’t’ represent the population therefore it may be inaccurate to generalise results to the wider population. A small sample usually doesn’t represent all the key interest groups within a population.

Validity and reliability of the methods

Reliability refers to the consistency and stability of the results obtained from a research study. If results are to be considered reliable then they should be very alike each time the experiment is repeated. If there are differences in important characteristics of the participants when the study is repeated or if the conditions under which the study was conducted are different then results cannot be considered reliable.

Eg in studies done at school, it may be difficult to replicate a study exactly as different students will do the study the next time, they have different gender mix, different personalities, the study may be done at a different time of day, or different time of year. It is difficult to conclude that these research studies are reliable.

Validity means that the research study has produced results that accurately measure the behaviour or event that it claims to have measured. Personality tests may not measure personality; they may just measure intelligence or ability to give socially desirable responses.

Another type of validity refers to the validity of the conclusions drawn by the researcher. Eg if a researcher concludes that a new drug reduces headaches, when it only reduces it for some people.

External validity refers to results being able to be valid outside a laboratory. Measuring sleep in a sleep lab, may not give the same when measuring sleep at home. Results collected in a lab must be externally valid in the real world.

Effect of extraneous variables – research may be considered invalid because one or more extraneous variables may not have been adequately controlled and have therefore influenced the results in an important way. Researchers cannot be certain whether it was the IV being tested or an unwanted extraneous variable that caused the change in the DV.

Descriptive Statistics

• Instruction in statistics should be limited to determining medians and means; generating graphical representations of data; interpreting medians, means, standard deviations, and graphical representations of data. However, a brief description of the function of inferential statistics and criteria for significance will enable students to read original research with some understanding.

Statistics are a vital component of any research they are used by psychologists to analyse and describe the dtate they gather, collate and summarise. The two types of statistics generally used by psychologists are descriptive and inferential statistics.

In order to describe large amounts of data in a meanigful way we identify patterns in the data. Data may be represented graphically or in numerical form.

Descriptive statistics look at 3 different characteristics of data:

1. central tendency (mean, mode and median)

2. distribution of scores

3. dispersion (how widely spread the data is – standard deviation)

Descriptive statistics are used for analysing, organizing, summarising and describing data collected.

These include measures of central tendency such as: mean (average), median (the middle value in a distribution) and mode (the most “commonly-occurring value in a set of data). Whilst most researchers prefer the mean (to include all data collected), the median is used when there are extreme or unusual values that could bias the mean.

Psychologists also attempt to summarise their data, by using frequency distribution tables, and graphs, such as line graphs, bar graphs, histograms, frequency polygons and pie graphs to simplify the data visually.

The standard deviation summarise how far on average a score differs from the mean. The standard deviation provides information about the variability or spread of scores in relation to the mean.

– A small standard deviation indicates there is little variability and that most scores are clustered around the mean

– The higher the standard deviation the greater the variability there is among the scores.

You need to be able to interpret a scenario such as this:

Consider two psych classes who do an exam out of 100. Both have a mean score of 65. One class has a standard deviation of two whereas the other has a standard deviation of 8. Conclusion; the class with the standard deviation of 8 may be harder to teach as the students vary more in ability. More scored well, but more scored poorly.

Ethical considerations

An overview of ethical considerations in the conduct of psychological research: role of the experimenter,

Before any psychological research is undertaken, the experimenter must make sure that moral principles and standards have been applied in order to maintain an appropriate level of participant care. The research needs to be carefully designed so that the data can be collected without harming the participants in any way, whether it is physical or psychological. Once the experimenter is satisfied that the study does not breach any ethical guidelines, he/she must submit the proposal of the research to an ethics committee who scrutinise it before giving the experimenter permission to proceed. During the course of the study, the experimenter must act professionally and not breach any ethical boundaries.

Participants’ rights (confidentiality, voluntary participation, withdrawal rights and informed consent),

Confidentiality refers to the principle that all personal information and responses of the participants in an experiment must not be revealed without their permission. Questionnaires or any data collecting devices should not display any of the participants’ personal details and, should the researcher need to retest in the course of a longitudinal study, information identifying the participants must be kept separate from their responses. Participants also have the right to participate anonymously.

Voluntary participation and Withdrawal rights,

In any psychological study, participants must be involved on a voluntary basis that is, they are given the freedom to decline participation. Participants must not be coerced, given inducements or threatened in any way. If they are given a payment it should only be to cover any costs incurred in being involved in the study rather than serving as a money making exercise. Participants should be told that they are free to withdraw at any stage of the study so their welfare is not compromised in any way.

Informed consent procedures,

Before any investigation begins, potential subjects need to decide whether to participate in an experiment after being informed of the nature and purpose of the research. As already stated, their participation must be totally voluntary and whenever possible, their consent should be obtained in writing.

Deception in research,

Often it is necessary to hide the true nature of an experiment and not openly discuss it with the participants before the data is collected. This approach attempts to prevent subject expectancy effects and obtain genuine, unbiased data. If subjects are led to believe that the researcher is investigating one particular question but is really investigating another, the subjects must be adequately debriefed after the data is collected.

Debriefing,

This is where the researcher, in post-experimental session, explains the true nature of the research to the participants. The participants are then given the opportunity to ask any questions about the research and to comment freely on any part of the experiment. This procedure ensures that the participants experience no lasting harm while still ensuring that the purpose of the investigation is not compromised.

Professional conduct

The researcher of any study should adhere to the ethical principles outlined above and should maintain a rigorous standard in each stage of the research. He/she should respect the rights of the participants and use the data collected to draw appropriate and valid conclusions in order to further the knowledge base in the given area. The objectives of the research should be clear and justify the treatment of the subjects, whether they are human or animal.

In Australia, the Australian Psychological society has a code of ethics that stipulates the standard of professional conduct, which researchers must abide by when conducting any study.

Four Levels of Explanation

Human behaviour can be observed and studied from different view points. This subject concentrates on four perspectives and refers to them as levels of explanation. You could say each level is like a different window.

|Biological |Basic Processes |

| | |

|Focus: The physical processes underlying |Focus: The psychological processes that are |

|behaviour - What is happening to the |common to most people |

|body | |

| |Involves: |

|Involves: |Memory |

|Brain structure & function |Learning |

|Sensory receptors & the motor system |Cognition |

|Hormones & Genetics | |

| |Example: An unrealistic fear of open spaces might be better understood through|

|Physiological features: |considering; |

|Changes in heart rate, breathing rate, |The original event - memory |

|blood pressure |What kind of learning may have taken place |

|Pupil dilation, perspiration |What thoughts are taking place - cognition |

| | |

|Example: Automatic arousal - ‘Fight or flight’ | |

|Person |Socio-cultural |

| | |

|Focus: The influence of individual differences on |Focus: The influence of other people on |

|behaviour |behaviour -The social context |

| | |

|Involves: |Involves: |

|Age, class, education |Culture |

|Personality |Socio-economic influences |

|preferences |Peer groups |

| |Stereo-types |

|Example: How comfortable an individual feels in the mosh pit at a rock concert| |

|may differ depending on; |Example: A males desire to cry in public may be influenced by; |

|music preference |Cultural norms in relation to the situation |

|age |The opinion and presence of peers |

|personality – introvert/extrovert |How he believes other perceive his tears |

Each topic is designed to concentrate on one level of explanation, with the final topic integrating all four levels.

|Topic |Level of Explanation |

|Social Cognition |Socio-cultural |

|Learning |Basic processes |

|Personality |Person |

|Psychobiology of Altered States of Awareness |Biological |

|Healthy Minds |Integration of levels |

-----------------------

Experimental treatment (IV)

Post-test

Pre-test

Experimental group

Post-test

Pre-test

Control group

Random allocation

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download