research methods 2

Primary data

Data collected by the researcher themselves for the purpose of the experiment

Strengths of primary data

Researcher has lots of control over the data

Studies can be designed specifically to collect particular data

Weaknesses of primary data

Lengthly and expensive

Possibly could be validity issues such as lack of ecological validity

Secondary data

Data collected for purposes other than the study

E.g government statistics, exam results, medical records

Strengths of secondary data

Data may not fit the needs of your study. The researcher has no control over the method of data collection uses. May not be valid or reliable

Weaknesses of secondary data

Data may not fit the needs of your study

The researcher has no control over the method of data collection used so may not be valid or reliable

Acronym for quantitative data

NOIR

N - nominal

O - ordinal

I - interval

R - ratio

Nominal data

Data in the form of categories

E.g. yes no

Ordinal data

Ranked data

Interval data

Data with real measurements with NO true zero

Ratio data

Data with real measurements with a true zero

Measures of central tendencies

Mean

Mode

Median

Mean strengths

Makes use of all the values and is therefore the most representative measure

Mean weaknesses

No good for nominal or ordinal data

Can be unrepresentative if there are extreme values

Mode strengths

Can be used with any type of data

Unaffected by extreme values

Mode weaknesses

No useful if there are various values that meet the criteria

Isnt useful for small sets of data

Median strengths

Unaffected by extreme values

Median weaknesses

Not as sensitive as not all values are reflected

No good for nominal data

Ignores very large or very small outliers

Range strengths

Easy to calculate

Range weaknesses

May be affected by extreme scores

Does not indicate how widely or tightly spread the data is

Does not use all pieces of data

Ways to measure spread of data

Range

Standard deviation

Standard deviation strengths

Much more precise measure of dispersion as it uses all the data

Easy to calculate (with calculator)

Standard deviation weaknesses

May hide characterises of the data set

E.g. any extreme scores

Can only be used with interval or ratio data

Ways of displaying data - tables

Uses raw data Or central tendencies

Ways of displaying data - bar chart

Display frequency data

Shows nominal data

Ways of displaying data - histogram

Y axis must start at 0 - no gaps between bars

Continuous data

Can be used with ratio data

Ways of displaying data - pie chart

Shows frequency data

Ways of displaying data - line graph

Used for continuous data

Ways of displaying data - scatter diagram

Correlation data

Use to calculate spearmanns rho

What does spearman’s rank test identify

Correlation coefficient

Which shows the strength and direction of the relationship between two variables

Normal distribution

Most of the examples in a set of data are close to the ‘average’ while relatively few examples tend to one extreme or the other

Normal distribution shaped like a bell - ‘bell curve’

Skewed distribution

Could have data set clusters around high or low values

Bell curve can be skewed in particular directions

Negatively skewed distribution

There are few extreme low scores and a skew towards highs scores

The mode is greater than the mean

Positively skewed distribution

Few extreme high scores and a skew towards low scores

A positive skew has a long tail on the right

Mean is greater than the mode

Inferential statistics

Tests use statistical calculations to state whether results are meaningful or just due to change

If results are not due to change they are SIGNIFICANT

Used to judge which hypothesis to accept or reject

Significance level

This refers to the minimum probability we will accept that our results are due to change

Usually aim for significance level of 5% - we are 95% certain our results are not due to chance

This is writtenm as P≤0.05

What is the number called that you get from a statistical test

Observed value

Tests where observed value had to be higher

Spearman’s rank

Chi squared

why do researchers use statstical tests

To determine whether the results in a study are statically significant

Three facts to consider when choosing a statstical test

D - difference - are you testing for difference or correlation

D - design - the experimental design - independent or repeated

D - data - what type you have - nominal or ordinal

What are the three D’s use for statistical tests

Difference

Design

Data

Useful way to remember statstical test table

Chi si chi

Britney willy spears

What is spearman’s rank and when to use it

Used when trying to find a correlation between two co-variables

Data is related to

Data collected has to ordinal, interval or ratio

How does spearmans rank work

Ranking two sets of data comparing the rank of each set

Then using it to find the relationship between them

The closer the rank of the two sets of data the stronger the relationship