Analysis Methods
Analysis Methods
Exporting data
If you are using an online survey program such as KoboCollect Toolbox, you will probably download
your file for further analysis. You can export data in various formats. Consider which format will be
easiest for you to work with based on the program you will use to analyze the data.
This is a special kind of distribution that large amounts of naturally occurring continuous data often
approximates. The normal distribution is sometimes known as the ‘Bell’ or ‘Gaussian’ curve. It has two key
properties:
1. the mean, median and mode are all equal
2. fixed proportions of the data lie within certain standard deviations (SD) of the mean: 68% within one SD,
95% within two SDs and 99.7% within 3 SDs.
MEAN AND MEDIAN
DISTRIBUTIONS
Means and medians are the most fundamental parameters of a numerical dataset.
The mean, or average, is the sum of all values over the number of measurements (n).
If a data set is sorted, the median is the mid value, i.e. where half of the measurements are larger or
smaller than this value.
For a large dataset comparing the means and medians gives an indication that the
data is normally distributed if the average and median are similar or even identical.
For a small dataset you can use statistical procedures to assess if a distribution
follows the required normal distribution before you can apply statistical methods for
data analysis.
When data does not follow a normal distribution
pattern
In almost all natural systems, parameters are
normally distributed. If you find they are not, there
are generally two reasons:
1. Outliers are skewing the data
2. You need to transform the data so that it will
follow a normal distribution.
PARAMETRIC AND NON-PARAMETRIC TESTS
1: Parametric tests - follow a normal distribution, and use the mean and standard
deviations in their calculations. Common parametric tests you may use in your
hypothesis testing include:
One sample t test
Paired samples t test
Independent samples t test
One-way ANOVA or multi factorial ANOVA
Correlation or regression analysis
2: Non-parametric tests - don’t require any continuous variables to be normally
distributed. Non-parametric tests use the ordinal position of data instead of the mean
and standard deviation in their calculations.
One common non-parametric test you may come across is the Mann-Whitney U test. This may be used
as an alternative to the independent samples t test if your data is not normally distributed.
HYPOTHESIS TESTING IN PRACTICE: ONE SAMPLE
T TEST
A one sample t test is used to test whether the sample mean of a continuous variable
is significantly different to a ‘test value’ (some hypothesised value).
For example: field or on-station trials where you are comparing a treatment with a
control.
HYPOTHESIS TESTING IN PRACTICE:
PAIRED SAMPLES T TEST
A paired samples t test is
used to test whether there is a
significant difference between
sample means for continuous
variables for two related
groups.
HYPOTHESIS TESTING IN PRACTICE: INDEPENDENT
SAMPLES T TEST
An independent samples test is used to test whether there is a significant difference in
sample means for a continuous variable for two independent groups.
For example: field or on-station trials with a paired comparison design, where you
are comparing the difference between two treatments in different research sites.
HYPOTHESIS TESTING IN PRACTICE: ONE-WAY
ANOVA
One-way or multi factorial
ANOVA is similar to the
independent samples t test,
but is used when three or
more groups are compared.
Data collected from quantitative research activities is numerical and collected under
controlled conditions for statistical analysis.
REPORTING AND COMMUNICATING QUANTITATIVE
RESEARCH
When communicating and reporting on the findings of this research,
quantitative researchers must decide what formats are most suitable to
communicate and present results. This will depend on the type of data
collected, and the research question that researchers are aiming to answer
with the data. For example:
Field observations and measurements, and survey data will usually be suited to
descriptive statistics and formats that describe the data visually using frequency
tables, bar charts, pie charts, line graphs, histograms, box plots, scatter plot
diagrams etc.
Experiments using inferential statistics to test hypothesis statements will report
statistical values such as P values, 𝝰-probability values, and confidence levels that
indicate the level of confidence in the result of the experiment. The values reported
by quantitative researchers will vary depending on the type of statistical test that is
used to test the hypothesis.
COMMUNICATING RESEARCH: KNOW YOUR
TARGET AUDIENCE
Knowing your target audience and the story of why you are communicating to them helps to
decide the format or methods that will be most suitable to communicate with them.
For example: the way you communicate internally to your fellow field researchers will usually
be quite different to the way you communicate with external stakeholders, or your senior
leaders.
Think about the audiences that you will be communicating your research data or results to. For
example, will you be communicating with:
Other field researchers?
Senior Researchers?
Project Leader?
Consider how you might need to adapt the format and way you communicate to suit different
audiences.
Communication etiquette refers to the accepted ways of communicating with others.
Good etiquette helps exchange information effectively while maintaining positive
relationships.
Decide on level of formality - adjust your tone and language to suit the audience and purpose of
communication.
Be respectful - listen, acknowledge others, and always offer constructive criticism.
Consider the timing of your communication.
Be concise but thorough- proofread all correspondence and reports.
Use a suitable format to deliver the communication.
Conclusion
This lesson has introduced you to data visualisation methods, and visual storytelling
techniques that you can use to communicate the findings, outcomes and impact of both
qualitative and quantitative data and research.
You've also considered the importance of knowing your target audience, and
adjusting your communication formats and methods to suit the audience - and
maximise the impact of your research communication.