PSYC2012

What is empiricism

reliance on systematic evidence

What is a hypothesis?

generalised claim about the world, testable statement

What is a prediction?

precise, accounting for the details of content, expectation of the study, operation and result.

Theory

Set of ideas intended to explain facts or events, broad, existing findings, underlying mechanisms

Claim

assertion that something is true

What is a variable and it's role in psychology?

anything that varies, allowing for measurement, manipulation and relationships between behaviour, cognition and emotion.

What is a dependent variable?

outcome or response

What is a independent variable?

Potential cause for the dependent variable

What is a construct?

mental processes, behaviours or traits that cannot be directly observed

Operationalisation

procedures designed to represent a construct

What are the 4 scales of measurement and their definition?

1. Nominal - Categories only

Data are grouped into labels with no order.

Example: Eye colour, gender.

2. Ordinal - Ordered categories

Categories have a ranking, but the differences between them are not equal.

Example: Satisfaction ratings (poor, fair, good).

3. Interval - Equal intervals, no true zero

Differences between values are meaningful, but zero does not represent the absence of the quantity.

Example: Temperature in °C.

4. Ratio - Equal intervals with a true zero

Has all the properties of interval data, plus a meaningful zero point.

Example: Height, weight, age, reaction time.

What is reliability?

degree of stability of measurement outputs across time or context

Types of reliability

Internal Consistency - Consistency among items measuring the same construct.

Test-Retest Reliability - Stability of scores over time.

Inter-Rater Reliability - Agreement between different raters or judges.

What is validity?

the degree to which a claim is correct?

Psychometric validity definition

attributes exist, variations in the attribute casually produce variation in the measurement outcomes

What are the types of vakidity and there definitions

1. Convergent Validity - The degree to which a measure correlates highly with other measures of the same or related constructs.

2. Divergent (Discriminant) - Validity The degree to which a measure has low correlations with measures of unrelated constructs.

3. Criterion Validity - The degree to which a measure correlates with an external criterion or outcome.

4. Concurrent Validity - A type of criterion validity in which a measure correlates with another measure assessed at the same time.

5. Predictive Validity - A type of criterion validity in which a measure predicts future outcomes or behaviour.

6. Content Validity - The degree to which a measure adequately represents all aspects of the construct or content domain being assessed.

7. Construct Validity - The degree to which a measure accurately assesses the theoretical construct it is intended to measure

What are the different types of patterns?

1. Bell shaped

2. Skewed (left, right)

3. Unfiform ( the bars are =)

4. Bimodal (2 peaks)

What is central tendency?

what value best represents a typical vaue/centre of distribution

Measures of central tendency

1. Mode - score with the highest frequency

2. Median - score the divides the distribution in 2 equal parts

3. Mean - average

4. Range - difference between largest ad smallest value

5. IQR - difference between the upper and lower quartile

Variance

Considers all values

deviation score =

score - meam

What does squaring each deviation score do?

means that all values are positive, providing mor e weight to larger deviations. - outliers affect variance more

Variance =

mean squared deviation - add all sqaured deviations together and divide by the number of scores

Deviation =

to obtain the original units - take the squareroot

Relative scoring

interpret a score with respect to it's relation to the mean

Standardization

converting raw data into a z-score

How to get a z-score?

Divide the deviation scores by the standard deviation

- how many standard deviatrions a value is

Proporties if the stanard normal distrubution

1. mean = 0

2. DS = 1

3. Total area = 1

4. mean, median, mode = nearly the same

Population

o all cases with the target characteristic

Sample:

o subset of the population

o Purpose of statistical methods is to allow us to confidently generalise from our samples to the population

methodological processes involved in testing a research hypothesis.

o Start with the assumption that there is no effect/association = null hypothesis.

o Seeking evidence against the null hypothesis

o H0 = Null hypothesis

1. Devise the intervention: operationalise the independent variable

2. How to assess the dependent variable: operationalise the dependent variable

3. Determine how to judge whether the intervention was effective: select a comparator-effective compared to what?

4. Collect data from people who have completed the intervention

5. Run a statistical decision

6. Draw a conclusion

After attaching a probability to results, if they are unlikely, what are the two possibilities?

(1) The null hypothesis is true and our results are unusual

(2) The null hypothesis is false

o The probability we use is the P value

Why try to find evidence against a null hypothesis rather than evidence for a research hypothesis?

falsifiability (stronger than confirmation)

What s a P-value

o Probability of obtaining to observed results if the null hypothesis is true

o If this probability is small = reject the hypothesis

o Not small, we retain the null hypothesis

o Small means <.05

Variance of the sampling distribution = σ ²/x̄ = σ² / n

σ²x̄ = variance of the sampling distribution of the mean

σ² = population variance

n = sample size

Standard deviation of the sampling distribution of the mean =

σx̄ = squareroot σ 2/x = σ 2/x /N - standard error of the mean

what does standard error of the mean do

- This gives us an indication of how much (on average) we expect each sample mean to vary from another sample mean of the same sample size (N)

Shape-the Central Limit Theorem

as the sample size becomes large, the sampling distribution of the mean will have an approximately normal (bell-shaped) distribution, regardless of the shape of the original population distribution.

Critical x̄ approach

determine whether to reject the null hypothesis by comparing the sample mean to critical values.

One tailed vs two-tailed test

One tailed = directional HA

Two tailed = non-directional HA

Types of qualitative data analysis

1. Thematic - themes/patterns

2. Grounded theory

3. Interpretative phenological analysis- how individuals experience and make meaning of it

4. Narrative - personal stories

Data collection methods

Semi-structures interview - open-ended

Structured interview - set questions, limited range response

Unstructured interview- opening statement

In depth interview

Focus groups - group interview

Observational - enthrograghy

What do t-tests do?

Determine whether differences in means are statistically signficant

One-sample t-test

Compare the mean of one sample to a known/hypothesised mean

Related-samples t-test

sam participants measured twice. mean difference between 2 mesurements

Independent samples t-test

2 unrelated group means comapred

Results of a t-test

1. t-value- size + direction of difference

2. degrees of freedom (n observations - n estimations)

3. p-value

Common effect size (Cohen's d)

A measure of effect size that indicates how large/meaningful the difference between two means is

- how much of the variance in the outcome can be explained by group differences

0. 2 = small

0.5 = medium

0.8 = large

What are confidence intervals and their role in statistical inference?

range of values likely to contain the population mean. 95% = common

If the CI does not contain the null value, the results = significant

Counterbalancing

varying the order of conditions across participants to prevent order effects from influencing results - imrpoves interval validity

Problem with repeated measures design

results can be influenced by order effects- better practise, fatigue etc.

When is one-way analysis of variance appropriate?

1. 1 IV with multiple levels

2. 1 continuous DV

3. Testing for overall group differences

What do the results of one-way analysis of variance tell us

1. F-test

2. Does not tell you which groups differ

What is the Mean squared within?

An estimate of population variance based on the combined influence of treatment effects and sampling variability on the group means.

eta squares (n2)

How much variance is explained across all groups (3+)

n2 = SSB/SST

SST = total sum of squares)

SSB = sum of squares between groups

Internal validity

The extent a study accurately estabkishes a causal relation between IV and DV. - threatened by confounding variables

External validity

ability to apply findings beyond the conditions of research

Differences between One-Way ANOVA and Two-Way

One Way =

effects of a single IV on the DV

Two Way =

effect of 2 IV's, their interaction and one DV

Accounts for more variability in DV

Interaction =

effect of one IV depends on the other IV vise versa

How to calculate components of an ANOVA table?

Factor A - main effect of IV 1