CODEWITHSUNDEEP
rtime-series
Nav
Nav

Reputation: 20668

R language: What does the result of the augmented Dickey Fuller test mean?

I've seen this, this and this, but none explain properly, the interpretation of the basic output that R shows. Also, some of the other questions about Dickey Fuller test didn't even receive answers, so I'm asking here.

Eg: From this tutorial, it says the null hypothesis is rejected, but I don't see anything in the output that indicates that. If I'm supposed to take a hint from the p value, then how do I know what is the cutoff limit considered for p? 

count_d1 = diff(deseasonal_cnt, differences = 1) 
plot(count_d1)
adf.test(count_d1, alternative = "stationary")

Augmented Dickey-Fuller Test

data: count_d1 Dickey-Fuller = -9.9255, Lag order = 8, p-value = 0.01 alternative hypothesis: stationary

In other tutorials I've seen such outputs where they conclude that the time series is stationary or non-stationary by looking at the adt.test output, but they mention nothing about how they came to that conclusion.

So how do I know if the null hypothesis was rejected or not? Isn't it possible to have an if statement like:
if adf.test(count_d1, alternative = "stationary")==TRUE, print("null hypo true"); else print("null hypo rejected");?

Upvotes: 1

Views: 2686

Answers (1)

Martin
Martin

Reputation: 332

Have a look at Walter Enders Applied Econometric Time Series 3e, 2010 for the theory. There is no strict rule of an "cut-off value". Nevertheless, a value of 0.05 is generally considered appropriate to discard a null hypothesis. (Of course, lower p-values support this thesis all the more credibly)

Upvotes: 1

Related Questions