Embedded Files
This standard is worth 4 credits and is about using 'bootstrapping' techniques to make inferences about populations we are interested in.
Here is a step by step guide to the processses you need to take to learn this at your pace!
Here is a step by step guide to the processses you need to take to learn this at your pace!
- know what inference topic is about
- explore the context/background story of your chosen data set for a practice assessment
- Pose your comparison question
- use data set of your chosen data to produce a pair of comparitive box/dot plots as initial analysis step
- write comparison paragraph between the two group
- conduct boot strap analysis on iNZight
- Discuss bootstrapping observations to assist you in making an informal inference
- Discuss sampling variabililty
- Write a conclusion and evaluation paragraphs
- consult the exemplars, NCEA Standard, Marking Rubrics and writing frames report structure to ensure that you know what you need to do to get success in your final report!
- ask your teacher to assign your formal assessment and follow steps 2-9 above again for REAL!
Step 1: An Introduction to Inference
Step 1: An Introduction to Inference
1 page summary of the Inference topic
framework for 3.10.pdfOverview of Formal Inference Report Writing (with exemplar paragraphs)
Formal Inference Report Writing.pdf
Step 2: Practice Datasets to learn through
Step 2: Practice Datasets to learn through
Important note: these datasets are CSV files - press "download as" then "comma separated values.CSV" and save them to your computer into a folder you can find, then open file in iNZight software or iNZight lite online.
Choose your theme for your practice assessment!
Choose your theme for your practice assessment!
Use the attached documents to read and understand where and what the data is showing you.
FILTERING if 3 or more groups
Don't forget, after your initial analysis in iNZight, if you want to investigate (& bootstrap) the difference between just 2 groups of data where the qualitative variable has more levels, or if you want to restrict your sample to just one type of group for a more meaningful comparison (eg just sports players etc), then you will need to FILTER by that level of the categorical variable, then Export and Save as a new csv file before opening in the VIT module of iNZight. (see videos below on how to do this:)
VIDEOS
Part 1: Using iNZight lite to create graphs and filter to only 2 groups before bootstrapping in VIT module
Part 2: Using iNZight to filter by numeric variable (to make more sense)
Step 3: posing your question
Step 3: posing your question
Powerpoint 1: Introduction to Inference
PPT 1 Introduction to Inference .pdf
Step 4: Produce dot and Box plots using iNZight lite
Step 4: Produce dot and Box plots using iNZight lite
Step 5: Write comparison analysis between the two groups
Step 5: Write comparison analysis between the two groups
Powerpoint 2: Analysis
This is similar to level 2 Inference analysis but with more context and research to explain what you are seeing in your box and whisker graph.
If you are looking at a variable that has more than 2 levels, you will need to choose which 2 groups you want to compare to see if there is a real difference in the population and then you will need to FILTER and EXPORT a smaller dataset to a csv before the next step using VIT module.
PPT 2 Inference Analysis.pdfStep 6: Conduct bootstrap analysis in iNZight lite
Step 6: Conduct bootstrap analysis in iNZight lite
IF YOU DIDN'T DO THE FILTERING IN STEP 2 YOU WILL NEED TO DO THIS NOW!
Don't forget, after your initial analysis in iNZight, if you want to investigate (& bootstrap) the difference between just 2 groups of data where the qualitative variable has more levels, or if you want to restrict your sample to just one type of group for a more meaningful comparison (eg just sports players etc), then you will need to FILTER by that level of the categorical variable, then Export and Save as a new csv file before opening in the VIT module of iNZight. (see videos below on how to do this:)
Use these videos to filter into ONLY 2 groups before you create your bootstrapped graphs:
Use these videos to filter into ONLY 2 groups before you create your bootstrapped graphs:
Part 1: Using iNZight lite to create graphs and filter to only 2 groups before bootstrapping in VIT module
Filtering data in iNZight lite to use in VIT.webmPart 2: Using iNZight to filter by numeric variable (to make more sense)
Step 7: Write about your bootstrapped graphs
Step 7: Write about your bootstrapped graphs
Powerpoint 3: Bootstrapping
You need to use the VIT module of iNZight.
Bootstrapping is a more sophisticated way of determining if there is a difference in the population based on re-sampling from within the original sample (assuming it is a reliable sample to start with).
This MUST be correct with a correct inference to pass the standard - make sure you understand what the confidence interval shows.
PPT 3 Bootstrapping Process updated.pdfVideos explaining how to conduct and interpret a bootstrap result
Videos explaining how to conduct and interpret a bootstrap result
Using Inzight to filter and subset data to export for inference.mp4Using VIT to bootstrap and interpreting the output.
Bootstrapping to make an inference of basketballers vs netballers reaction times.webmBootstrapping using online VIT module.webmLink to VIT: https://www.stat.auckland.ac.nz/~wild/VITonline/
Step 8: Write about sampling variation
Step 8: Write about sampling variation
Powerpoint 4:Sampling Variability - a very important concept to make sure you discuss clearly.
PPT 4 Sampling variability.pdfStep 9: Writing Conclusions and Evaluations
Step 9: Writing Conclusions and Evaluations
3.10 Theory Notes based on Sigma.pdfVideo from Stats @ UofA - Confidence Intervals and the Bootstrapping process
Powerpoint 5: Conclusion
Answering your question! Summarising your investigation and evaluating the process, as well as giving ideas for further analysis.
PPT 5 Conclusion and Evaluation.pdfPowerpoint 6: Some ideas about how to show excellence level thinking - go beyond the first thought/analysis - go deeper!
PPT 6 Aiming for excellence.pdfStep 10: Exemplars, rubrics and Assessment criteria
Step 10: Exemplars, rubrics and Assessment criteria
Note: these are exemplars only, and do not necessarily provide you with a guide that you should follow, more ideas of how concepts can be discussed in one context only - you need to be able to explain your understanding in the context of your assessment dataset. To gain a good mark, you would be unlikely to merely follow an exemplar as your analysis would be specific to your findings.
Report abuse