Research Methods

?

Experimental Method

Aim: A general expression of what the researcher intends to investigate.

Hypothesis: A statement of what the researcher believes to be true. It must be operationalised, or measurable. There are two forms of hypothesis, a directional hypothesis has a clear direction. E.G x will increase. Non-directional do not have a direction, E.G x will change.

Anull hypothesis is essential in a study, and states that nothing will change. E.G x will not change

Experimental method: A researcher causes the IV to vary and records the effect of the IV on the DV. E.G X affects Y in a certain way, but if we change X how will it change the effect on Y?

1 of 26

Research issues

Extraneous and Confounding variables:

  • Extraneous Variables are 'nuisance' variables that do not vary systematically with the IV. A researcher may control some of these
  • Confounding Variables change systematically with the IV so we cannot be sure if any observed change is due to the CV or IV. CV's must be controlled. An example is age in a weight gain study, as age could affect the results of the experiment.

Demand characteristics refer to any indication of the true aims of the study which could affect how the participants react to the experiment and possibly skew the results and make them anomalies. 

Investigator effects are when the researcher gives accidental cues to the participants, giving them hints on how they should behave which can negatively impact the results of the experiment. The researcher is usually unaware of this. 

2 of 26

Research Techniques

Randomisation: Using change to randomise the designs of the investigations to control the effect of bias. Examples include randomising the amount of times a condition is repeated. 

Standardisation: Using exactly the same procedures for each participant in the study

Control Groups: Control groups are really important in studies, and they are established to use a group to compare the effects of the study. For example, in a study looking at the effects of coffee on women, a control group would consist of women doing tasks who havent drunk any coffee. 

Single and double blind: 

  • Single blind is a research method in which the participant doesnt know the aims of the study. To stop ethical issues arising, the participant is told the aims of the study after the experiment in a debrief after the study. To stop soley demand characteristics. 
  • Double blind is a research method in which both the participants and the researchers dont know the aims of the study, to stop both investigator effects and demand characteristics. 
3 of 26

Independent Groups

An independent group consists of two groups, one group doing condition A and one doing condition B completely seperate of eachother. They are randomly allocated. This experiment type has several advantages and disadavantages. For example:

+There are no order effects, the participant only has one chance to do the action in the experiment so they wont get bored or gain practice from continually doing the same task. 

+The participants will not guess the aim, as again they are only tested once so it is unlikely they will guess the aim of the study. So, in theory it would be hard to show demand characteristics. Therefore, behaviour should be more natural

-Despite this, participant variables mean that the study can become less valid as individual differences can possible act as confounding variables and affect the results of the study

-Also, as the researcher has to gather double the amount of participants for the two groups it is more expensive as more time spent is recruiting and more money must be paid out. 

Overall, independant measures studies are more effective at limiting order effects and demand characteristics, but are more expensive and dont take into account individual differences. 

4 of 26

Repeated Measures

Some participants will take part in all conditions of an experiment. The order of conditions should be counterbalanced to counter order effects. 

+Participant variables are controlled, as the participants in both conditions have both characteristics. This controls individual differences, an important CV.

+Fewer participants means more time studying instead of recruiting

-Order effects are a major problem, as the participants are doing a similar task twice. This reduces the validity of the results, as the participants may do better or worse on the second attempt. 

-Quite opposite to independent groups, participants may change their behaviours as they guess the aim of the study. This will also reduce validity.

Overall, while the study is cheaper and participants variables are controlled, the overall validity of the study is reduced by order effects and demand characteristics. 

5 of 26

Matched pairs

Two groups are assembled where the participants have as similar characteristics as possible. So, because of this participant variables are reduced. Both groups have a condition to do like independant. 

+Participant variables are reduced, as people are matched based on similarties between them.

+There are no order effects, as participants are only tested once. So, there are no fatigue of practice effects and this overall increases the validity of the study. 

-Matching isn't perfect, and participants unless identical twins wont be literally identical. So there will be participant variables but they will be reduced.

-Need twice as many participants, which is more expensive and time consuming

Overall, much like independant groups matched pairs has no order effects and they wont guess the aim. However it improves upon independant variables, as there are less paticipant variables. However, matching isn't perfect and you need double the amount of participants. 

6 of 26

Types of Experiments

Types of Experiments

NAME ME ONE NOW

                                                             

7 of 26

Labratory Experiment.

A controlled experiment where extraneous and confounding variables can be regulated. Participants will go to the researcher, and the IV is manipulated and the effect on the DV is recorded.

+EV's and CV's can be controlled and their effects on the DV can be minimised. Cause and effect between the DV and IV can be differentiated and demonstrated. So, it has a high internal validity. 

+Lab experiments can be easily replicated, as the procedure to create the experiment can be replicated as it is highly standarised. If the results are the same this confirms their validity. 

-However, studies in a lab setting may lack generalisability as they are conducted in such an artifical setting, which can make the participants aware of the fact that they're in a study. Thus, behaviour might not be natural and so it cant be generalised to everyday life. Therefore it lacks external validity.

-Also, because of the artificial setting the participants may start to understand the aim of the study which can lead to demand characteristics.

Overall, Lab studies are more controlled and replicable. However, the artifical setting creates problems. 

8 of 26

Field experiment

Field experiments are conducted in natural settings, and the participants still go to the researcher. The IV is manipulated and the effect on the DV is recorded.

+As it is set in a more natural environment, participants are more comfortable and therefore the results will be more generalisable to everyday situations.

+Participants are unaware of being studied however, and while this could have ethical issues they are more likely to behave as normal if they are unaware they are being studied. The study has greater external validity. 

-It is more difficult to control CV's however, as observed changes in the DV may not be due to the IV but CV's instead. So, due to this it may be more difficult to establish cause and effect. 

-There are also many more ethical issues, as the participant may not give informed consent in a field experiment. This is an invasion of participants privacy. 

Overall, a field experiment is more generalisable to the outside world, and as they are unaware of being studied they act more natural. However, CV's are harder to control and there are several ethical rules such as informed consent being breached.

9 of 26

Natural Experiments

The researcher does not manipulate the IV. The IV would have varied even if the experimenter wasn't interested. DV may be naturally occuring or may be measured by the experimenter. 

+It may be the only ethical option, as it may be unethical to manipulate the IV. E.G studying the effects of children in care. So for some topics it is the only ethical way. 

+Greater external validity, as natural experiments involve real life issues. For example how a natural disaster affects stress levels. This means the findings are more relevant to experiences.

-The natural event may occur rarely, so this reduces the oppurtunity for research and this may limit the scope for generalising findings.

-Participants are not randomly allocated, so CV's are completely uncontrolled and this can affect the results dramatically.

Overall, while natural experiments are very extrenally valid and ethical, they are hard to conduct and CV's are a major issue concerning the results.  

10 of 26

Quasi-Experiment

A quasi experiment has an IV based on a pre existing difference, betwen people, for example gender, age or social status. No one has manipulated this variable, it just simply exists. DV may be naturally occuring, or amy be measured by the experimenter. 

+There is often high control in Quasi experiments, as they are similar to lab experiments when conducted and so share many benefits with lab experiments. 

+Comparisons can be made about people, as in a quasi experiment the IV is the difference between people. For example, people with or without autism. 

-Participants are not randomly allocated, which means the experimenter has no control over which participants are placed in which condition. So, participant variables may have caused change in the DV acting as a CV

-Casual relationships are not demonstrated as the researcher does not manipulate the IV, so we cannot say for certain that any change in the IV affected the DV.

Overall, quasi experiments measure the set differences between people. They are highly controlled, but suffer from the uncertainty of the IV actually affecting anything. 

11 of 26

Population and Sampling

Population: A large group of people that a researcher is interested in studying, for example college students from North West

Sample: Because a population is usually so large, it is very hard to include them all in a study. So, reseachers rely on samples from the larger population to represent it. 

Generalisation: This sample as it is from a parent population can be generalised to it, as the participants in the sample are from the same population as everyone else.

Bias: However, the majority of samples are biased in that certain groups may be over or under represented. A stratified sample however would represent all people in a population

12 of 26

Opportunity sample

Most avaliable:  People who are most avaliable i.e the ones who are nearest/easiest to obtain

How: Ask people nearby e.g ask the students in your class to take part or people in a shopping centre. 

+Quick method: Opportunity sampling is convienant because you just make use of the people who are close to you. Therefore, it makes it the most popular sampling method 

-Inevitably Biased: The sample is unrepresentative of the target population as it is drawn from a very specific area, such as one street or town. This doesnt represent the entire population, and this means that finding cannot be generalised. Also, a researcher may avoid people based on appearance as they look for people personally to join their study. 

13 of 26

Volunteer Sample

Self-Selecting: In a volunteer sample, participants select themselves 

How?: Advertise: For example, place an ad in a newspaper or on a noticeboard and paticipants come to you

+Participants are wiling: Participants have selected themselves and know how much time and effort is involved, and so they will be more engaged in the terms of the study compared to people from an oppurtunity sample. 

-Likely to be a biased sample: Participants may share certain traits, for example being keen or curious. Due to this participant variable, generalisation will be limited due to this bias. 

14 of 26

Random Sample

Equal chance: Every person in the target population has an equal chance of being selected 

How?: Lottery method. All members have an equal chance of being selected for the study with a tombola or a hat picking method. 

+Potentially Unbiased:  The researcher has no influence over who is selected so the study is free from researcher bias. So, they wont choose people who support their hypothesis. 

-Representation not guaranteed: Despite being random, its entirely possible that the results of the random sample are completely unrepresentative of the larger population. You might've gotten the 5 richest people in the poorest population. 

-Time consuming: It is very time consuming and very difficult to obtain data for everyone. Participants may also refuse to participate. 

15 of 26

Systematic Sample

Sampling frame: Participants are selected using a set 'pattern'

How?: Every nth person is selected from a list of the target population

+Unbiased: The first item is usually selected at random and because of this systematic sampling has quite an objective method. 

+Fairly Representative: The sampling method represents the population pretty well. 

-Time and effort: A complete list of the population is required, and with the amount of time it makes you could just as well use a random sampling. 

16 of 26

Stratified sample

Frequency: Participants are selected according to their frequency in the target population

How: Subgroups or strata are identified, such as gender or age groups. The relative percentages of the subgroups in the population are reflected in the sample. 

+Wholly representative: the characteristics of the target population are represented as the sampling method takes into account all of the characteristics of the population. Therefore, it is the best generalised to the whole population.

-It is not perfect: Strata cannot always reflect all the ways in which people are different. So, unfortunately complete representation is not possible. 

17 of 26

OI

GIVE ME AN ADVANTAGE OF VOLUNTEER SAMPLING 

                                                                                                                                

18 of 26

Ethical Issues

Ethical Issues

Conflict - When an issue exists betwen the rights of the participant and the aims of the research.

Informed Consent - Participants should be able to make an informed judgement about if they want to take part. 

Deception - Deliberately misleading or withholding information so consent is not informed

Protection from harm - Participants should be at no more risk than they would in everyday life. Also, they should be given the right to withdraw at any time and should provide counselling if there was harm.

Privacy/Confidentiality - We have a right to control information about ourselves, if this is invaded confidentialiality should be respected.

19 of 26

Correlation

Correlation means association

A scattergram is used to measure correlation. One axis has one condition, and the other axis has the other condition. Correlation can go one of three ways, Positive, Negative or Zero. Positive meaning that the co-variables rise or fall togethe, negative meaning one co variable rises or falls. None meaning no correlation. Basically, positive is up, negative is down and zero is all over the place. In an experiment the researcher will manipulate the IV and records the effect on the DV. In a correlation there is no manipulation of variables and so cause and effect cannot be demonstrated. 

+ Useful starting point assess the strength and direction of a relationship can be used for future hypothesis

+Economical No need for controlled enviroment, no manipulation needed, explains relationship

-No cause and effect cause and effect are not properly distuinguished may be intervening variables. 

- Method possibly flawed can be flawed and not reliable, reducing the validity of the study.

20 of 26

Observation

Observational techniques 

A way of seeing or listening to what people do without having to ask them. Often used to assess the DV.

+Can capture unexpected behaviour - People often act differently from how they say they will in self report methods, and observations are useful as they give an insight into spontaneous behaviours.

- Risk of observer bias - Researchers interpretation of the situation may be affected by expectations. Bias can be reduced by having more than one observer. 

21 of 26

Types of observation and evaluations.

Naturalistic - Takes place where the target behaviour would normally occur.

+ High external validity - In a natural context, behaviour is more likely to be spontaneous. Therefore, naturalistic observations are more applicable to real life. 

- Lack of control - However, despite this as the setting is natural there could be uncontrolled EV's which could affect the results and make pattern detection poor.

Controlled - some control/manipulation of variables including control of EV's

+ Can be replicated - More easily repeated due to standardised procedures, so findings can be found again via replication. 

- Possible low external validity - Behaviour may be contrived from the setting - demand characteristics may be present. So, these findings cannot be generalised to everyday life.

22 of 26

Even more observational techniques! Overt and Cove

Covert - participants are unaware of being studied.

+ Demand characteristics reduced - Participants do not know they are being watched, so their behavioural will be more natural. Increases validity of the study. 

- Ethically questionable. - People may not want their behaviour recorded in public. Right to privacy possible affected. 

Overt - Aware of being studied

+ More ethically accepted - Participants have given their consent to be studied, so can withdraw at any time.

- Demand characteristics - Knowledge of being studied influences behaviour, reducing the validity of findings. 

23 of 26

Participant observation

Participant - when the researcher becomes part of the group they are studying

+Can lead to greater insight - Researcher experiences the situation as the participants do, enhancing the validity of the study. 

-Loss of objectivity - The researcher may identify too strongly with those they are studying - going native. 

Non-participant - when the researcher remain seperate from the group they are studying 

+ More objective - Researcher maintains an objective distance so less chance of bias. Smiling and grinning 

- Loss of insight - Researcher may be too far removed from those they are studying. May reduce the validity of findings. 

24 of 26

Self report techniques

Questionnaires

Questionnaires are made up of a pre set list of written questions to which a participant responds. They can be used as part of an experiment to assess the DV.

+ Can be distributed to lots of people, which allows for data to be gathered quickly and without the presence of the researcher. Reduces the effort involved and makes questionnaires cost effective. 

+ Respondents may be willing to 'open up', so they may share more personal information as they aren't under scrutiny. So, there is less chance of social desirability bias. 

- Responses may not be truthful, as participants will want to present themselves in a positive light so social desirability bias is still possible.

- Response bias is an issue, as respondants favour a paticular kind of response e.g they always agree 

25 of 26

Interviews

Interviews are face to face between an interviewer and interviewee 

Structured Interviews 

+Easy to replicate Straightforward to replicate because of standardised format. This format also reduces differences between interviewers

- Interviewees cannot elaborate and cannot deviate from their points this may be source of frustation. 

Unstructured Interview 

26 of 26

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »