Time Series Regression and Prediction Assignment

Time Series Regression and Prediction Assignment 1. Collect Data Set(s) 2. Examine the data. Discuss BRIEFLY. Might there be a fundamental change in the series (Hint: think about the global warming discussion & 1963 ). What other issues show up? 3. Determine relationships & format data set a) what is your dependent variable – format your data set appropriately b) explain why you consider this a time series data set c) do you think there might be seasonality to the trend? If so, why? How will you handle the seasonality in your modeling approach? 4. Do some correlation analysis a) Do a scatter plot, or multiple scatter plots if needed b) Does the evidence suggest that there is a time trend? If not a trend does it show some other time series pattern (auto-correlation? Moving average?) c) Evaluate your time span. Is it too short? Is it too long? Should you look only at a smaller portion to determine trend? If it looked as if a fundamental change occurred, adjust your data set accordingly. d) Add a fitted line to the portion you’ve chosen. Do not add the formula at this point. Check - a linear trend - quadratic trend - an exponential trend - OR, moving average (if necessary) Which trend seems to fit your time series best? e) Write out the equation of your model based on 3(a),3(c),4(d) – NOT THE FITTED equation, just the generalized specification 5. Run a regression on data set a) Do your estimated co-efficients and intercept make sense? Check to make sure you used “t” instead of year or some other factor. b) write out your estimated model b) NOW ADD the fitted line equation to your scatter plot. c) Compare your estimated model to the one generated by Excel (or mini-tab). Are they consistent? (Hint: make sure you’re comparing apples to apples) 6. Fitted equation a) interpret your model (i.e. what do the estimated coefficients & intercept mean)? b) predict the next three values in your time series. Comment on these predictions (i.e. do they make intuitive sense? What do they mean? c) COMMENT ON THE REGRESSION. What else might you do? Are you satisfied with the model? Does it lack something? If so, what might that be? Are there possible non-independence issues that you’d like to resolve? Etc. Maybe look at the residuals. Do they show some additional pattern that you’d like to explore?