0

I'm reading through a paper on economic factors and how they pertain to elections. Unfortunately I don't remember enough stats to fully understand how to read the tables being presented. Here is a sample...

"Data are aggregate time series gathered over the post-World War II period. The percentage of the public approving of how the president is handling his job, according to a Gallup poll, is the variable to be explained. The independent variables beside the economy are war, political scandal, international crisis and term cycles. Examples of such a model appear in Table 1..."

enter image description here

At first I thought the values being represented were the number of percentage points each variable was responsible for. But -306% doesn't seem to make sense in that regard? I'm also unclear what the "Term dummies" category is for.

My questions are --

1) If possible, I'd love to get a primer on what the data in this table represents

and

2) A pointer towards what type of stats study I should do to better understand what is going on here.

Thanks!

1 Answers1

1

As it written on the table explanation at the bottom, first column uses ARIMA (Autoregressive Integrated Moving Average) which is a model that suits for time - series analysis.
This model seems a bit complex and you can't simply interpret the coefficients as you do in your good ol' regression model, try this: ARIMA model interpretation for a lead.

The second and third columns use OLS which is linear regression, there's plenty of stuff on it on the web, but shortly this model assume a linear connection, i.e:
$y = \beta_0 + \beta_1x_1+\beta_2x_2+...+\beta_px_p+\epsilon$

The numbers in the column are the estimated coefficients ($\beta_i$) meaning the effect of $x_i$ on $y$ while holding all other constant. The asterisk near them indicate whether they are statistically significant (loosely speaking - it's plausible these numbers are not caused by pure chance).

If you don't understand the meaning of the features (the $x_i$) like Term dummies, I advise you to open the original paper that this data was drawn from.

RiskyMaor
  • 112
  • 1
  • 1
  • 9