Message from Nikephoros in Vibrant Diversity #general


2017-04-11 21:29:06 UTC  

Do you have a TA that's of any use?

2017-04-11 21:29:20 UTC  

Do you know anyone who's already taken the class and could help?

2017-04-11 21:29:22 UTC  

No TAs

2017-04-11 21:29:28 UTC  

Rip

2017-04-11 21:29:37 UTC  

TAs are usually useless anyway

2017-04-11 21:30:02 UTC  

From what I understand, It's very straightforward if you know what these outputs mean

2017-04-11 21:31:17 UTC  

Like does Prob > F = (#) mean Fail to reject or reject H0

2017-04-11 21:33:40 UTC  

@Nikephoros I have no idea, I haven't used R since I used python

2017-04-11 21:33:55 UTC  

Which I haven't done since python 2.7 came out

2017-04-11 21:34:14 UTC  

I know a guy who's premed who could help but I'm not sure I'll be able to get a hold of him

2017-04-11 21:34:29 UTC  

I have a friend at Case Western, I can shoot him a text and ask if he knows

2017-04-11 21:34:36 UTC  

So none of those outputs mean anything to you?

2017-04-11 21:35:25 UTC  

Lol not at the moment

2017-04-11 21:35:41 UTC  

Sorry, but I'll ask a couple of friends and see if they can help

2017-04-11 21:36:46 UTC  

ok the simplest way you can ask them is this: "Do these results indicate that I can reject H0?"

https://cdn.discordapp.com/attachments/274262571367006208/301470658779676683/Screenshot_20170411-173535.jpg

2017-04-11 21:44:14 UTC  

You know what, I think the decision rule is Reject H0 if "Prob > F" is greater than "F(#, #)"

2017-04-11 21:44:29 UTC  

Which appears to be true

2017-04-11 21:45:05 UTC  

https://cdn.discordapp.com/attachments/274262571367006208/301472751787442186/Screenshot_2017-04-11-17-44-22-1.png

2017-04-11 21:45:54 UTC  

https://cdn.discordapp.com/attachments/274262571367006208/301472956406824961/Screenshot_2017-04-11-17-45-33-1.png

2017-04-11 21:46:06 UTC  

@Koba Tell him this is the output to a Granger Causality Test

2017-04-11 21:46:11 UTC  

I did

2017-04-11 21:46:26 UTC  

He's a bit of an autiste so he takes a while to get his thoughts out

2017-04-11 21:48:28 UTC  

I see. I'm fairly certain that I can reject H0. Even if I don't know how to explain it using the data, the underlying question is: Is "egg" the cause of "Chicken"

2017-04-11 21:49:02 UTC  

The answer is, presumably, yes.

2017-04-11 21:52:45 UTC  

@Nikephoros I wish I could offer help. I used STATA for econometrics as well, but that was nearly 10 years ago.

2017-04-11 21:53:34 UTC  

I did my econometrics project on whether baseball player salaries are representative of objective output.

2017-04-11 21:54:04 UTC  

https://cdn.discordapp.com/attachments/274262571367006208/301475012869881856/Screenshot_2017-04-11-17-53-27-1.png

2017-04-11 21:54:08 UTC  

Spoiler alert, they aren't unless you time shift the pay by like 2 years IIRC.

2017-04-11 21:57:41 UTC  

@Nikephoros is the > an operator or just an output format?

2017-04-11 21:57:54 UTC  

@Koba I really don't know why its formatted "Prob > F", I think it might refer to a critical value of some sort.

2017-04-11 21:57:59 UTC  

Maybe not

2017-04-11 21:58:27 UTC  

Because I read that as "probability of f = .8903"

2017-04-11 21:58:58 UTC  

"Prob of chi2 = .8877"

2017-04-11 21:59:00 UTC  

There is nothing special about granger causality versus other hypothesis tests in stata.

2017-04-11 22:00:01 UTC  

So what should I interpret these outputs to mean?

2017-04-11 22:00:16 UTC  

Reject H0, right?

2017-04-11 22:00:39 UTC  

https://cdn.discordapp.com/attachments/274262571367006208/301476670718738433/Screenshot_2017-04-11-18-00-15-1.png

2017-04-11 22:01:01 UTC  

So, I read your results as failure to reject

2017-04-11 22:01:30 UTC  

Ah so I was reading it backwards

2017-04-11 22:01:37 UTC  

Yes, I agree with that. I think you have to accept H0 with those numbers because it doesn't meet the confidence interval.