Research methods 2 Psych

quantitative data

numerical data that can be measured and analysed statistically allowing objective comparisons

qualitative data

descriptive data expressed in words providing depth and insight into behaviour

primary data

data collected first hand by the researcher for the specific study increasing relevance and control

secondary data

data collected by someone else and used by the researcher providing convenience but less control

nominal data

data in categories without numerical value allowing simple frequency counts

ordinal data

data that can be ranked but intervals between values are not equal limiting statistical precision

interval data

data with equal intervals between values but no true zero allowing more precise analysis

ratio data

data with equal intervals and a true zero allowing the most powerful statistical tests

mean

the arithmetic average of a data set providing a sensitive measure of central tendency

mean strength

uses all data values making it representative

mean limitation

affected by extreme scores reducing accuracy

median

the middle value in an ordered data set providing a measure unaffected by extremes

median strength

not influenced by outliers increasing accuracy in skewed data

median limitation

less sensitive because it ignores most data values

mode

the most frequently occurring value providing a simple measure of central tendency

mode strength

useful for categorical data

mode limitation

may be unrepresentative if multiple modes exist

range

the difference between the highest and lowest values showing overall spread

range strength

easy to calculate

range limitation

affected by extreme scores reducing accuracy

standard deviation

a measure of how far scores deviate from the mean showing consistency of data

standard deviation strength

uses all data values giving a precise measure of spread

standard deviation limitation

affected by extreme scores

normal distribution

a symmetrical bell shaped distribution where most scores cluster around the mean

skewed distribution

an asymmetrical distribution where scores cluster at one end affecting measures of central tendency

positive skew

most scores are low with a long tail to the right shifting the mean above the median

negative skew

most scores are high with a long tail to the left shifting the mean below the median

correlation

a statistical relationship between two variables showing how they change together

positive correlation

as one variable increases the other also increases

negative correlation

as one variable increases the other decreases

zero correlation

no relationship between the variables

correlation strength

useful for identifying relationships for further study

correlation limitation

cannot establish cause and effect

scattergram

a graph showing the relationship between two variables used to identify correlations

bar chart

a graph showing discrete data using separate bars making comparisons easy

histogram

a graph showing continuous data with touching bars representing frequency distribution

x-axis : range of values or bins of the data,

y-axis : represents the frequency or number of occurrences

line graph

a graph showing changes over time using connected points

pie chart

a circular chart showing proportions of categories

content analysis coding

categorising qualitative data into meaningful units allowing quantification

thematic analysis

identifying recurring themes in qualitative data providing deeper understanding

raw data

unprocessed data collected directly from participants before analysis

descriptive statistics

numerical summaries of data such as mean and standard deviation

inferential statistics

tests used to determine whether results are due to chance

levels of measurement importance

determines which statistical test is appropriate ensuring valid conclusions

frequency table

a table showing how often each value occurs organising data for analysis

percentage

a proportion expressed out of 100 allowing easy comparison between groups

measure of central tendency

a value representing the centre of a data set such as mean median or mode

measure of dispersion

a value showing the spread of data such as range or standard deviation

inferential statistics

statistical tests used to determine whether results are due to chance allowing researchers to draw conclusions about populations

significance level

the probability threshold for rejecting the null hypothesis usually set at 0.05 to balance accuracy and risk of error

critical value

the value that the test statistic must exceed to reject the null hypothesis ensuring decisions are based on probability

observed value

the calculated test statistic from the data compared against the critical value to determine significance

type one error

rejecting the null hypothesis when it is actually true meaning a false positive conclusion is made

type two error

failing to reject the null hypothesis when it is actually false meaning a false negative conclusion is made

parametric test

a statistical test requiring interval data normal distribution and equal variances providing more power when assumptions are met e.g T test

non parametric test

a statistical test used when data does not meet parametric assumptions allowing analysis of ordinal or non normal data e.g Chi squared

chi square test

significant association between two categorical (nominal) variables.

compares observed frequencies and expected frequencies

chi square requirement

expected frequencies must be above five to ensure accuracy of the test

sign test

a test used for repeated measures and nominal data to assess differences in direction of change

to see whether the difference between two means is significant.

sign test calculation

count positive and negative signs and compare the smaller value to the critical value

wilcoxon test

a test used for repeated measures and ordinal or interval data to assess differences between related scores

non‑parametric statistical test used when your data is ordinal or not normally distributed.

wilcoxon requirement

data must be at least ordinal to allow ranking

mann whitney test

a test used for independent groups and ordinal or interval data to assess differences between two samples

non‑parametric test used to compare two independent groups

mann whitney requirement

data must be at least ordinal and samples must be independent

spearman rho

a correlation test used for ordinal or interval data to assess the strength and direction of a relationship

spearman rho requirement

data must be at least ordinal and variables must be monotonic (moves in one direction)

pearson correlation

a correlation test used for interval data and normally distributed variables to assess linear relationships

pearson requirement

data must be interval and show a linear relationship

related t test

a parametric test used for repeated measures and interval data to assess differences between related means

related t test requirement

data must be interval and normally distributed

unrelated t test

a parametric test used for independent groups and interval data to assess differences between two means

unrelated t test requirement

data must be interval normally distributed and have equal variances

anova

a parametric test used to compare means across three or more groups to identify overall differences

anova requirement

data must be interval normally distributed and have equal variances

degrees of freedom

a value based on sample size used to determine the critical value for statistical tests

one tailed test

a test predicting a specific direction increasing power but increasing risk of type one error

two tailed test

a test predicting a difference in either direction reducing risk of type one error

effect size

a measure of the strength of a relationship or difference providing information beyond significance

power of a test

the probability of correctly rejecting a false null hypothesis increased by larger samples and stronger effects

pilot statistical check

using small scale data to ensure the chosen test is appropriate before full analysis

reporting significance

stating whether the observed value exceeded the critical value and whether the null hypothesis was rejected

interpreting non significant results

concluding that evidence was insufficient to reject the null hypothesis but not proving it true