Data Analysis & presentation

Presentation & Interpretation of quantitative data

The raw numerical data must be presented, analysed and interpreted in order to make sense of them. This can be visually in tables, graphs, bar charts, histograms & scattergrams & they can also be analysed statistically.

Graphs

• Graphs acts as visual aids that help make sense of quantitative data.
• Provide an overall picture that helps to summarise the results.

A line graph is often used to indicate changes over time. E.g. data relating to no. of student enrolled every year for the AS ps exam over a 10 year period in the U.K

It is important to ensure that continuous data only are presented in line graphs to prevent inappropriate interferences being drawn.

It is also important to be aware that manipulation of the way that the axes in graphs are drawn can easily bias the interpretation.

Conventions that facilitate the presentation of info in an unbiased way & reduce  the risk of misunderstanding/ misinterpretation.

• Plot frequency of scores on the y-axis
• Adopt the three-quarter-high rule:this states that when frequencies are plotted, the length of the y-axis should be determined in the following way. It should be presented in such a way that the distance of the highest point on the graph from the x-axis is approximately equal to 3/4 of the total length of the x-axis
• Need to clearly label, with a clear & informative title

Bar Charts

• Usually used when the data are in categories/ to illustrate the average scores from different samples.

Histograms

• Useful form of graphical representation that can be used when presenting continuous data on interval/ ratio scales of measurement.

Frequency Polygons

• Useful technique when it is necessary to compare 2 or more frequency distributions. It can be used instead of a histogram.

Scattergrams

• one way of presenting correlational data is a scattergram
  

Tables

• Frequency table is more useful & helps to make any patterns within data more obvious.
• Important that the columns & rows clearly labelled.
• When a large no. of different values exist for a variable, presenting the data in a frequency table can become complex. instead the different values can be organised in a small no. of grouped values & presented in a grouped frequency table - it is important that the intervals used are mutually exclusive, exhaustive & of equal width.

Analysis & interpretation of quantitative data

Data from observational studies can have threats e.g.inadequate sampling & mishandling of the system used.

Categorization is important tasks, it groups items together

The median- mid value of scores arranged in ascending/ descending order

• remains relatively unaffected by any outlying values, therefore it is safe to use when unsure of reliability of extreme values or data from skewed distributions.  (cluster at one end of range)
• does not work well with small data sets & is affected by any alteration of the central values in a set of values.

The mean- arithmetic average

• makes use of all available data, most powerful measure, can be used for skewed data.
  
• when it has been calculated, decimal points may be less meaningful if all data has consistent numbers.
  

The mode- most frequent

• easy to identify
• even v.small changes in data can radically alter result, may not be single number,
  

Range-  diff between high and low.

• quick to calculate
• does not provide any idea of the distribution of values around centre, nor does it take individual values into account, if any outlying values BAD!!

Interquartile range- measure spread of mid 50% when in numerical order (rank ordering)  so use mean than 255 upper boundary and 25% lower is lower boundary

Standard deviation- measure of variability/ spread of given set of scored from its mean,

postive - as one variable increases so does other

negative - as one variable increases other decreases

Systematic research technique for analysing transcripts for interviews, documents, ads, cartoons, films, lyrics, texts (visual/ written), newspapers, websitesd

The researcher creates a coding system of predetermined categories at the outset of the study, which is then used on the selected materials in a transparent & consistent manner.

A pilot study is carried out to generate & test the coding system to be used. care is needed to ensure the categories are discrete & do not overlap.

Coding qualitative data consits of assigning specific elments of the material to be analysed to a specific category 7 then drawing interfences about the themses relfected in the content. The requency of occurence of specific works is sometimes analysed, in this way the reseracher can count how many time each code (word) occurs, therby extracting quantitative data from qualitative in objective way.

Can get 2nd person to check coding, or code data independently & compare outcome.