Data handling and analysis

Quantitative vs qualitative data

1. Quantitative data: involves measuring something, statistical analysis, collected in experiment based research methods

Strengths: objective and scientific, less prone to researcher bias. Data collection methods are efficient

Weaknesses: lack of depth, not capturing complexities. Reductionism and simplifying

2. Qualitative data: involves finding out what people think I'm more detail, collected in case studies.

Strengths: Rich, in depth understanding, detailed insights. Contextual understanding, holistic views

Weaknesses: subjectivity issues, researcher bias. Time consuming and resource intensive

Primary vs secondary data

Primary: collected directly from participants by researchers for their own research aim. Obtain first hand knowledge to test a hypothesis. Sources are questionnaires, interviews, observation

Strength: authentic data, controlled extraneous variables

Weaknesses: time and resources, needs planning and prep.

Secondary: collected or created by someone for their own purposes, researcher applies and recycles it to their own hypothesis. Can be for meta analysis. Sources are articles, websites or books, meta analysis

Strengths: easy access, inexpensive and not as time consuming

Weaknesses: loss of extraneous variable control. Data not obtained for the specific purpose of your investigation

Meta analysis

A collection of different research studies, independent. Trends and patterns. Re- analyze existing dataset. Statistical

Strengths: high statistical power, resolves conflict with refuting research. High generalisability and high sample

Weaknesses: “garbage in garbage out” problem. Eg. Bad studies cause a bad meta analysis. Selection bias

Measures of quantitative data

1. Measures of central tendency- info about the typical score (averages). Reduces a large amount of data (raw data) to a single value eg. A number, to represent that entire data set. Three measures: mean, median, mode

2. Measures of dispersion- info about how spread out the scores are (variability).

Mean

1. Mean: the statistical average set of data, calculated by adding all the scores together and dividing by the number of values.

Strengths- uses all the scores so is powerful and sensitive. Weakness- can be distorted by extreme scores that are much higher or lower than the rest, aka. Outliers/anomalies

2. Median: middle central value of all datanafter ranking in ascending order.

Strengths- unaffected y extreme scores, easy calculations. Weakness- only takes account of middle values.

3. Mode: score that occurs most often. Calculated by frequency count

Strength- unaffected by extreme values, easy calculation. Weaknesses: not useful in small sets of data, doesn't account for other values

Measures of dispersion

1. Range- subtract highest from lowest

Strengths- quick and easy calculation, takes account of extreme scores. Weaknesses- distorted by very extreme scores, fails to a show if scores are evenly spread around the mean

2. Standard deviation- sophisticated method of spread.

Strength- takes account of all scores and is sensitive measure of spread. Weakness- less meaningful if data isn't normally distributed, time consuming

Calculating sd

1. Calculate difference between each score and the mean

2. Square each of those differences

3. Add squared differences together

4. Divide by total no. of scores in data set

Percentages

%: Number/total x100

% decrease or increase : difference/original x100

Presenting data

1. Bar charts- non continuous (nominal) data

2. Histograms- continuous scale of measurements, shows distribution

3. Pie charts- circular statistical data

4. Scattergrams- correlations and relationships

5. Summary/results table- summarise main findings

Distributions

1. Normal distribution: any given attribute or behaviour will gain a score that centres on the mean. Symmetrical and the median mode and mean are all at the peak of the curve.

2. Negatively skewed data: more high scores than low scores. Ceiling effect as data is clustered around the high scores. Test is easy. Mean is lower than the mode

3. Positively skewed data: more low scores than high scores. Floor effect because data is clustered around low scores. Test is hard. Mean is higher than the mode

Levels of measurements

Nominal data: categories eg. Colours, 20-40 and 40-60. Frequency count, factors only belong to one category.

Ordinal data: numbers placed in ascending or descending rank order. Has a true zero. Eg. Pain levels from 1-10. True intervals are unknown between each individuals score (my rating of 7/10 isn’t the same as someone else’s rating of 7/10)

Interval data: gaps between each interval are equal and set and known. Eg. The difference between 100-70 are the same units as the difference between 80-60

Nominal, ordinal and interval

1. Nominal: mode

Strengths: simple, clear eg. tally’s. weaknesses: lack of depth only mode

2. Ordinal: mode, median

Strengths: useful for complex, subjective data. Weaknesses: gaps in between are unrankable

3. Interval: mode, median, mean

Strengths: can be converted to ordinal data, more sensitive testing. Weaknesses: reductionist when comes to complex ideas, lacks true zero