Sin título

Página 1

PS2010: Research & Methods

There are 3 types of psychological research: Descriptive, Experimental & Correlational

There are 2 types of research methods used in psychology: qualitative and quantitative.

Quantitative research uses observational and

Qualitative research:

Collects data concern with meaning & experiences.

How people make sense of the world, understanding phenomena

Exploring question rather than investigate variables & hypotheses

Data is naturalistic- direct from participants e.g. words & images

Inductive (bottom up) not Deductive (top down)

Subjective meanings not objective meaning.

E.g : Participant meaning & elements of researchers' interpretation

Can generate theory or gain deeper insight

Research designs:

- Participants observation, interviews, dairies, questionnaires, focus groups

- Usually small numbers of participants

- Same ethical consideration

Other methods used

Newspaper articles, Historical records, blogs, facebook & tweets

Methods of Analysis

Data analyse range of ways

Linked to research question

Method should influence way data is collected & who participants are

Common methods of analysis

Content analysis - categorising info in data less qualitative

Thematic analysis - foundsational method, identify themesfrom data

Content analysis – categorising information in the data,

less qualitative

Thematic analysis – foundational method, identifies

themes which emerge from the data

Grounded theory – identifies themes used to derive

theories, data collection continues until emerging

themes ‘exhausted’

Discourse analysis – interested in the use of language,

features of speech within data, how meaning is

constructed

Interpretative phenomenological analysis –

understanding experience of particular group of people

Independence measures –ANOVA (Analysis of variance)

One-way Independent measures ANOVA

measure one variable (more than one condition)

-Comparing 2 or more

groups

- Multiple t test or ANOVA (takes into account all

conditions)

-Experimental variance: due to experimental

manipulation

-variability between conditions correct

-Random variance: measure error, individual

difference, unmeasured variables

- Variability within conditions X

Use F ratio

to explain random variance (Exp Var

/ Random Var)

Large F ratios the greater the variance =

significant results

E.g. Ps look at pictures and are given a memory

task

-IV1: time to inspect scene (4 levels: 5secs,

10secs, 15secs, 20secs)

-DV: Rating number of objects recalled out of 10

SPSS Output:

Descriptive statistics- Mean & SD

Assumption: Levene’s test: if not significant,

homogeneity has been met

e.g. (F (3, 6) = .2, p=.916)

Main Effect:

-Better F ratio

-Anything below 1 is random junk

-Anything above 1 is good. Therefore, F > 1 mean Exp var larger than

Random var

-IV1 = Time

- F=MS time/MS error

- F= 83.5/ 1.2

-F=68.3

Write up: Main effect of time (F (3, 36) =68.3, p =.001).

The number of items recall differs significantly according to time allowed to

inspect the scene.

There was no significant difference between

conditions.

Bonferroni t tests

-

Compare conditions

-

Need 6 t test

-

Use Familywise error: 6*5 % Type 1 error

-

SPSS output:

If Levene’s test is not significant homogeneity is

assumed

-

If Levene’s test is significant use select Games-Howell

if in doubt

Bonferroni correction:

- Change

alpha level (criteria for significance)

-

Bonferroni alpha = .050 / number of comparisons to be made

-

Bonferroni alpha = .050 / 6

- Bonferroni

alpha = .008

- Run all

pairwise comparisons with t tests

-BUT – t

tests are only significant if p ≤ .008

One-tailed hypothesis -Two-tailed

hypothesis

-Directional prediction -No directional

prediction

-Planned comparisons - Post hoc comparisons

-Planned

comparisons compare group conditions in different ways.

-Think

about specific and directional hypotheses.

- Interpret significant and direction

Planned comparison choices

1. Planned

comparisons: Deviation

-Compares

the effect of each condition (except the first – defined as “reference”) to the

overall effect

Contrast

one 2 vs. 1,2,3,4

Contrast

two 3 vs. 1,2,3,4

Contrast

three 4 vs. 1,2,3,4

2.

Planned comparisons:

Simple

-Compares

the effect of each condition to the control (defined as “reference” and

“first”) condition

Contrast

one 1 vs.2

Contrast

two 1 vs. 3

Contrast

three 1 vs. 4

3. Planned comparisons: Difference

-Compares

the effect of each condition to the overall effect of all previous conditions

4. Planned

comparisons: Helmert

- Compares

the effect of each condition to the overall effect of all following conditions

5. Planned comparisons: Repeated

- Compares

the effect of each condition to the next condition

Contrast

one 1 vs. 2

Contrast

two 2 vs. 3

Contrast

three 3 vs. 4

Think

about your research question and hypothesis.

-Want

to compare

different treatments to

a control? Simple

-Want to

see how

DV changes across

conditions? Repeated

Contrast

Estimate

Largest

estimate value for the trend that best describes the data.

Interpret

best fitting trend.

Writing up an ANOVA

•Which

type of ANOVA did you run and why?

-

What is the IV? Independent or repeated? Number of levels?

-

What is the dependent variable?

•How

do you break down the main effect?

-

Planned or post hoc tests? Which contrast and why? Hypothesis!

How

did you analyse the data?

•Give

statistics using APA standards and interpret the finding.

•F (df1,df2) = F ratio,

p = significance

Assumption

of homogeneity of variance met?

•Give

statistics using APA standards and interpret the finding.

•F (model df, error df) = F ratio, p = significance

Was

the main effect significant?

•Use

graph, descriptives and planned/post hoc contrasts

•Use

all three in combination to explain where differences are.

ANCOVA: Analysis

of Covariance

-A

confounding variable or control variable

-

Takes into account the variability cause by confound variable (use ANOVA model)

-

Same amount of experimental variance in percentage 1/3 or 33%

-Ratio

of experimental to random vary greatly

Stage1: is

covaraite significant, does it explain significant variability

Stage2: After taking covariate into

account is there more experimental than random variability

- If

more random than experimental variability ANOVA will not be significant (NS)

same for ANCOVA

I- If

explain lots of random variance covariate will be significant. Larger amount of

experimental variance significant.

SPSS

output:

New Page

New Page