# Research methods

HideShow resource information
• Created by: h
• Created on: 19-02-13 17:59

## LEVELS OF SIGNIFICANCE!

Levels of Significance: A SIGNIFICANT RESULT IS ONE WHICH IS UNLIKELY TO BE DUE TO CHANCE FACTORS.

(p=probability)(<, means, less than or euqal to )

p<0.5 (50% too lienient), p<0.01(1% too strict)

1. TYPE 1 ERROR: research hypothesis is accepted when infact it should be rejected p<0.5
2. TYPE 2 ERROR : research hypothesis is rejected when infact it should be accepted p<0.01
1 of 10

## INFERENTIAL STATISTICS!!

deciding whether or not to conduct a test :

• whether you are looking for a difference between conditions or a realtionship b/w variables
• what experimental design has been used
• what level of data has been used

nominal data ---------------------Chi Squared

correlational----------------------- Spearmans Rho

independent groups ---------------------Mann-whitney

Wilcoxon

2 of 10

## INFERENTIAL STATS 2 !!!

once a test has been selected:

• calculated value must be compared to critical value
• the calcualted walue must be greater than the critical value in a chi -squared and spearmans rho test for the results to be seen as significant.
• for the results to be significant when using the Mann-whitney U TEST or the Wilcoxon ranked test , the calculated value must be less than the critical value

a criitical value is a value that statisticians have estimated to represent a significant effect and depends upon the level of significance required

3 of 10

## INFERENTIAL STATS 3!!

3 factors to consider when finding the critical value:

• N = number of participants
• type of hypothesis
• significance level (0.05)
• The calcualted Value must be greater than the critical value in a chi -squared and spearmans rho test for the results to be seen as significant.
• for the results to be significant when using the Mann-whitney U TEST or the Wilcoxon ranked test , the calculated value must be less than the critical value

comparing the calculated value with the relevant table value allows the researcher to make a conclusion deciding whether or not they should accpt the research hypothesis and reject the null hypothesis or vice versa.

REMEMBER: A RESEASRCH HYPOTHESIS IS NEVER PROVEN. THE STATISTISCAL ANALYSIS ONLY PROVIDES SUPPORT FOR THE RESEARCH HYPOTHESIS.

4 of 10

## FEATURES OF SCIENCE!

OBEJECTIVITY- This means that the findings gatherd are factual and unaffected by beliefs,  opinions, feelings and expectations. Total objectivity is not possible( WHAT WE OBSERVE WILL BE INFLUENCED BY WHAT WE EXPECT TO SEE).

REPLICABILITY- Research findings need to be replicable so an imporatnat aspect of any scientific research is that researchers in a particular field can check each others findings. Confidence in reseach findings is increased when studies are repeated and findings are similar. In order for replication to be possible, all details of the original study must be published, icluding data, proceedures and results.

FALSIFIABILITY- This means that it should always be possible to prove a theory wrong i.e.it must have a testable hypothesis. Popper says that rather than finding evidence to support a theory, scientists should activelt try to find evidence to show that it is false.

PARADIGM- Kuhn states that science operates through paradigms. A paradigm is a general world view that is accepted by a majority of scientists working in that discipline e.g. behaviourism. Itdetermines how researchers carry out thier work and also what is accepted as appropriate research.

5 of 10

## VALIDATION:PEER REVIEW

What is peer review? IT IS A PROCESS OF VALIDATING RESAERCH.

• research is conducted
• psychologists send draft to journal company
• editor reviews it and sends copy to expert in field
• peer  reviews read drafts and consider all aspects of study
• appropriateness of method and samplic technique, controlled variables and potential bias, appropriate conclusions are all looked at in depth
• peer  provides detailed feedback and an indepedent jugemnt regarding publication.
6 of 10

## CONDUCTING RESEARCH.

TITLE: precise so that the readeer can get a good idea of what the research will be about.

ABSTARCT: brief intro inclusing design, participants, outline of results and future research

INTRO: why is the research bieng conducted?

HYPOTHESIS: null, experimental and operationalised.

METHOD: Design +justification, variables

Participants; number, smapling+justification, conditions and dropp outs

apparatus

Standardised procedure; step by step research of how experiment was conducted.

7 of 10

## ......continued

RESULTS:

• Descriptive stats; mesures of centeral tendencey/dispersion. results should be presented in a graphical form
• inferential stats; statistical test used , levels of significance and whether the test was on tailed or two tailed

DISCUSSION:

• explanation of findings; discuss the findings obtained.
• do the findings support hypothesis?
• what is the most important finding
• relationship to background research
• limitations and modifications
• further research?

CONCLUSION: recap key findings ans conclusions

8 of 10

## ......continued 2

REFERENCES: list of material reviewd and used

APPENDICIES instructions given to participants,

GENERAL CONSIDERATION AND STYLE;  researcher should write report in third sense e.g researcher hav conducted an experiment NOT i have conducted an experiment.

CONSENT FORM MUST BE INCLUDED!

9 of 10

## EXAM MARKS!!!!!

DESCRIBE AND DISCUSS FINDINGS (8 MARKS)

• DESCRIBE; describe what is shown in tables,graphs,descriptive statistics,significance levelsuse actual figures from the question.
• DISCUSS; Interpret the findings,what do the findings demonstrate and suggest.
• can include disscussion comments such as realisim of the  task, measuring what was intended to measure, whether or not the variables were operationalised.

DESIGN A STUDY(12MARKS)

• clear and well organised/reasones design with sufficeint hypothesis
• operationalised,eperimetal and null hypothesis
• method +justification
• IV, DV,EV and co-variables
• sampling method + size
• procedure
• consent form- date, debrief and ethical guidelines(such as DRIP) must be added
10 of 10