WJEC A2 Psychology PY3 - Research Methods

RESEARCH/ALTERNATIVE HYPOTHESIS vs  NULL HYPOTHESIS

either:

EXPERIMENTAL or CORRELATIONAL

and:

DIRECTIONAL or NON-DIRECTIONAL

Research/Alternative: a statement predicting what you would expect to happen in a study which is clear and testable, often starts "that pps who", should contain the word "significant"

Experimental: a statement predicting how the IV will affect the DV

Correlational: a statement predicting how one covariable correlates with another covariable

Null: a statement predicting every possible outcome in a piece of research other than what you expect to happen; goal of research is to reject null hypothesis and accept research/alt

To turn a hypothesis into a null hypothesis:

Older people learn significantly slower than young people =

Older people do not learn significantly slower than young people, any change in learning speed is due to chance factors alone

A hypothesis can be Directional or Non-Directional (1 or 2 tailed):

Directional: statement predicting change/relationship and the direction of change/relationship

eg children who eat breakfast show significant increase in concentration compared to those who don't

Non-Directional: statement predicting the change/relationship of the DV but not the direction of change or relationship of the DV

eg a drug affects a person's ability to concentrate

Important: data refers to the actual scores of each individual pps, NOT the mean, median, total etc of group of pps

In order of sophistication:

Nominal = qualitative or quantitative

Ordinal = qualitative or quantitative

Interval = quantitative only

Ratio = quantitative only

Nominal Data (Categorical): usually used to classify/categorise when collating data

• Quantitative = number data with no numerical value (but refers to a category) eg house numbers, football shirt numbers, numbers of buses, reg numbers
• Qualitative = non-number data which refers to a category eg favourite colours, gender, type of music, yes or no answers

Ordinal Data (ranked): relationship between data ie ranked but difference between ranks may be different

• Quantitative: number data with a numerical value BUT without equal intervals between each rank on the scale eg ranking people on height, positions in a race
• Qualitative: non-number data which indicates a particular order but difference between each category isn't equal eg small, medium, large, grades of exams, any rating scales on a qs

Interval Data:

Quantitative:  number data with a numerical value where the difference between points on a scale are equal and there is no absolute zero eg temperature in degrees - can have minus

Ratio Data:

Quantitative: number data with a numerical value where the difference between points on a scale are equal but with an absolute zero and no minus numbers eg temperature in Kelvins, length in cm, time in secs

Nominal Data:

(+) can work out the frequency of the categories and produce charts, work out the mode/percentages

(-) inappropriate to use the mean, median, range, SD

Ordinal Data:

(+) can work out the mode, median to summarise data/compare data

(-) inappropriate to use the mean

Interval and Ratio Data:

(+) can use mean, median, mode + SD to summarise/compare data

(-) not qualitative, so may lack rich descriptive detail

p < 0.05 = probability of the null hypothesis being true

Significance level: the level which you choose to reject your null hypothesis, always the 5% level or p < 0.05 (95% sure it was the IV that caused the DV or, with correlations, that the covariable 1 coincides with covariable 2, and 5% sure they were due to chance)

Types of Statistic tests:

• Chi²
• Sign
• Mann Whitney
• Wilcoxon
• Spearman's Rho

Choosing a stats test: 3Ds

• looking for a difference (experiment or correlation/relationship)
• if there is a difference, what design
• what type of data did you generate? ("nominal" or "at least ordinal")

Choosing a stats test:

OBSERVER VALUE (OV) = calculated value after the raw data has been processed by the stats test

CRITICAL VALUE (CV) = value you compare with the OV that appears in the back of textbooks in critical value tables