When writing you report, no matter what type of data you're investigating, you will need to follow the PPDAC cylce. PPDAC stands for Problem, Plan, Data, Analysis an Conclusion. Below is a guide for what is expected in each section of your report. It is split up into "the essentials" - things that must be in your report for you to pass, and the "ways to improve your grade" - things you can add to help get you to a Merit or Excellence.
Remember, your overal grade is also heavily influenced by how well your report is written - is it easy to read and follow? Do you use correct spelling and grammar? Do you use statistical terminology correctly? etc.
Check out these notes made by Mr. Bradly
The essentials
What are you investigating?
Give a hypothesis (guess the answer to your problem)
Ways to improve your grade
Explain why your results could be useful in a real life context
Who might find the results of your investigation useful, and why?
Back up your hypothesis with contextual ideas
The essentials
Describe how you're going to collect your data, including how much data you will collect
If applicable, describe any cleaning you have conducted on your data
State what types of graphs you will use, and what analysis you will do
Comparison Data: You will be using a side-by-side box plot, comparing average, spread and shape.
Relationship Data: You will be using a scatter plot, describing the relationship with justification using a Q-Plot, and making a prediction
State at least two sources of variation and how they can be managed
Click here for more information about sources of variation.
Downloading Data from Census At School:
The essentials
State where your data has been collected from
Describe the variables you are investigating and what units have been used to measure them
Give a link to the spreadsheet where your data is stored
Regrouping Data in Census At School:
The essentials
Create a side-by-side box plot using NZGrapher. Your graph needs:
Appropriate labels and titles
Summaries
High Box Plot
Shape Outline
DBM & OVS (Numbers)
Mean Dot
Make a comment about:
Average
What is the median for each group?
Which group is bigger/smaller and what's the difference?
Spread
What is the interquartile range?
Which group has more/less consistent data?
Shape
Is the data unimodal or bimodal?
Is the data left-skewed, right-skewed or symmetrical?
Are there any outliers?
Ways to improve your grade
Give contextual reasons for anything you have noticed in your analysis
Make the call if the difference is reflected in the population
Find the difference between medians (DBM)
Find the overall visible spread (OVS)
State which threshold you will use
Datasets of around 30: DBM needs to be greater than 1/3 of the OVS to make the call
Datasets of around 100: DBM needs to be greater than 1/5 of the OVS to make the call
Datasets of around 1000: DBM needs to be greater than 1/10 of the OVS to make the call
If DBM ÷ OVS is greater than 1/3 (or 1/5 depending on dataset size) then you can make the call.
The essentials
Create a scatter plot using NZGrapher. Your graph needs:
Appropriate labels and titles
Mean Dot
Quadrant Count
Show Ratio
Make a comment about:
Trend
Is the data linear, non-linear or has no relationship?
As variable 1 increases, what happens to variable 2?
Direction
Is the data positive or negative?
Back this up with evidence from the Quadrant Count Ratio
Strength
Is the data strong, moderate or weak?
Back this up with evidence from the Quadrant Count Ratio
Interesting or unusual features
Identify any interesting or unusal features by their variable entries
Explain what makes then interesting or unusual
Remove the Quadrant Count on your graph and add a Custome Line (Check "Cust. Line" and "Show Eqn"
Fit your line of best fit through your data, passing through the Mean Dot
State the equation of your line
Make a prediction using your line of best fit, giving a prediction interval
Ways to improve your grade
Give contextual reasons for anything you have noticed in your analysis.
Make a comment about any clusters or identify that there are no obvious clusters
Colour by categorical variables to see if you can find any hidden clusters and comment about them.
Relate any findings to the context
The essentials
Give an answer to your original problem
Describe whether your analysis matches up with your hypothesis?
Ways to improve your grade
Following this investigation, what could you investigate next?
What data could be collected to help improve this investigation?