# Step 6

# Data Analysis 2

# Organising Data

# You have collected your data but now you need to organise it. The three organising data option below allow you to organise your data collection into a more manageable format. This allows you to view your data in a more visual friendly way. For example, imagine you have collected 1,000 results from 1,000 different participants. Looking for distribution, trends, similarities, differences in 1,000 results would be impossible. The three organising data options below would allow you to appropriately arrange your 1,000 results.

## RANK ORDER DISTRIBUTION

You only use rank order distribution if you have less than 20 participants! To organise data in rank order distribution you simply place the participants scores in order from highest to lowest. As a result, rather than looking at 19 different scores you can instead view them in order. This allows you to see variance (differences), similarities easier. However, it is only useful if there are less than 20 participants scores.

## SIMPLE FREQUENCY DISTRIBUTION

You use a simple frequency distribution table if you have more than 20 participants. However, only if you have a small range scoring data. For example, a research project looking at how many basketball hoops are scored by each participant during 5 throws. This is a small range because it is a score of 0 to 5. You could easily have a table showing how many players scored 1 shot, 2 shots, 3 shots etc up to 5. If you had a study that looked at number of points scored per player over a whole season you would have a large range of scores, probably 0-100. It would be ridiculous to have a table identifying how many score 1 point, 2 point, 3 points, 4 points all the way to 100 points! If this was the case you would do a ‘grouped frequency distribution’ table.

Once you have created your distribution table you can then see how your data/scores are distributed. You can see if most people scored low, middle or high. If you were doing an experiment you could create a distribution table before and after a treatment and then see if the treatment changed the distribution of results.

## GROUPED FREQUENCY DISTRIBUTION

You use a grouped frequency distribution table if you have more than 20 participants. However, only if you have large range scoring data. ‘Grouped frequency’ involves grouping your data scores into manageable chunks. For example, if you had a study that looked at number of basketball points scored per player over a whole season you would group the results into 0-10, 11-20, 21-30 etc. You could then create a table that showed how many players scored between 0 and 10 points, how many between 11 and 20 points etc.

Once you have created your distribution table you can then see how your data/scores are distributed. You can see if most people scored low, middle or high. If you were doing an experiment you could create a distribution table before and after a treatment and then see if the treatment changed the distribution of results.

## DISTRIBUTION GRAPH

Distribution curves are a graph that provides a visual representation of a simple of grouped frequency distribution table. The graph allows for a good visualisation and presentation of your result's distribution.

You can display more than one group on your distribution curve graph to compare groups, for example a curve to show distribution of heights for football players, another curve to show distribution of heights for basketball players. You can then visually compare the differences.

In most scenarios you would expect to see a normal distribution, where most people get a middle score, with a few people getting low scores and a few people getting high scores. For example, you would expect most participants to run 30m between 4 and 5 seconds, a few people to run slower than 5 and a few to run faster than 4 seconds. However, sometimes you can get a negatively skewed curve (where most people have a low score) or a positively skewed curve (where most people have a high score).