Assignment
Prepare an analysis of some time series data of your choice. There is no need to prepare a report, just do working in Excel that you can use to present your findings. Use power-point for presentation too.
Variable selection and data collection
You should choose a single variable for which there is unbroken, evenly-spaced time series data available. The variable you choose should also be one that potentially will have a seasonal component. This normally means you will need data observations at a higher frequency than annual – quarterly, monthly, daily or hourly data could all have regular high and low “seasons”. If you choose an annual time series you should be able to justify why you think there would be a seasonal pattern – e.g. a variable under significant influence by the government, such as tax receipts, could potentially follow a seasonal pattern over the electoral cycle if governments are more likely to increase tax rates just after elections rather than just before them.
You can choose appropriate data from any database you can legally access and are allowed to publicly disseminate. Look at the example data for chapter 14 exercises for ideas of the kinds of data you might look for. There would be many straightforward variable-of-interest candidates amongst labour market measures, tourism data, and financial time series. Statistics New Zealand’s Infoshare tool (http://archive.stats.govt.nz/infoshare/) is one place where you can find some useful time series data. There is also high frequency (even daily) data related to Covid-19 available from the Statistics NZ home page.
Ideally you would have at least 30 time points of data, if more is available then you should use it, but don’t go overboard. E.g. we might be interested in monthly live births but even if data is available going back to the year 1950 we may not want to use all of it, we might just be interested in the trend over the last 20 years and so only include the data going back that far.
Time series analysis and prediction
Your goal is to select the most appropriate model for your time series and use it to make predictions for the series two time-steps forward. For instance, if you are modelling a quarterly time series and the last observation was 2019 Q2 then you would predict for 2019 Q3 and 2019 Q4.
The models you should select from are:
· A smoothed trend curve computed with the Holt-Winters method – choose the parameters for the method you think give the most representative trend series
· Autoregressive models up to an appropriate lag length, depending on the likely seasonal period. E.g. for quarterly data you would estimate AR(1) to AR(4) models.
· Linear, quadratic and exponential trend models – you should estimate each of these without and with seasonal dummies. The number of dummy variables to add would depend on the likely seasonal period. E.g. for monthly data you would need eleven dummy variables.
From the above models, choose the one you think is best to use to make predictions for the next two time-steps. You should mainly consider model explanatory power, the significance of model coefficients, specification problems revealed by residuals, and the principle of parsimony in making your decision.
Then, use your preferred model to predict the value of the time series for the next two time-steps.
Presentation
You will present your findings in Power point and use excel for workings. Assessment criteria
The assessment criteria I will use is as follows:
· Content: have the models noted above all been sufficiently considered?
· Delivery: have the variable selection, data collection, time series analysis and predictions all been explained adequately?
· Discussion: does the discussion concerning the model selection indicate an informed decision?
Producing and using Powerpoint slides will be useful and impressive for you during the presentation, so you are welcome to use some slides.
For your reference, I provide the slides that the were used above. Bear in mind that student switched between the slides and their workbook during their presentation. I am unable to provide you with their workbook. The student clearly had put a huge amount of work in, beyond expectations, and delivered a professional presentation. So it is a challenging assignment, but very doable with careful management.
Refection paragraph
After the presentation, please individually submit a paragraph reflecting on the group work. The sorts of things I would like you to reflect on are:
· Were you happy with how tasks were allocated within the group? E.g. did the workload turn out to be fairly balanced?
· In what ways did the group work well together as a team? E.g. did leadership help to keep everyone motivated and on track?
· What were some aspects of working in the group that were challenging? E.g. were there any cross-cultural or interpersonal issues?
· What is something significant that you learnt while working with the group? E.g. did you learn that a way of dealing with a problem was successful or unsuccessful?
· Is there anything you would do differently next time? E.g. can you see where a communication breakdown occurred and how it might be avoided next time?
WE HAVE DONE THIS ASSIGNMENT BEFORE, WE CAN ALSO DO IT FOR YOU
GET SOLUTION FOR THIS ASSIGNMENT, Get Impressive Scores in Your Class
CLICK HERE TO MAKE YOUR ORDER on Statistics New Zealand’s Infoshare tool