turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- SAS STATS Questions

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-23-2016 08:53 AM

Dear SAS Experts,

Being a novice to the statistics, I would like to receive the answers for following questions in layman's term to rememeber it for long term.

1. Why we need to reject the null hypothesis when alpha value is less than 0.05?

2. What are the various ways to select the variables for regression?

3. How to find whether the model has fit our data? I remember we need to look for R Squared value and P value significance in the output. What other I should look for model fitness?

Thanks in advance for any help you offer me.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-23-2016 10:11 AM

1. That's not correct, 0.05 is commonly used but it's not required. Your level of significance depends on the application. In drug testing it may be 0.001.

Your other questioms are too broad. Have you taken the SAS Statisitcs 1 course? Or any of the statistics classes with Coursera? You may find them helpful.

2. What type of regression?

3. What type of model? What's your criteria.

Statistics is is not an exact science, it's about margins of errors.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-23-2016 01:48 PM

I agree Reeza.

1. I would like to know why we need to reject the null hypothesis when it is not significance? Appreciate if you could tell me in layman's terms.

2.Linear Regression and Logistic Regression

3.Liner and Logistic model. What we need to look in output apart from R squared and parameter estimate?

I've also a other question

4. How to figure out the multicollinearlity between the indepedent variables used in the model?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-23-2016 02:24 PM

You're asking for a course in a paragraph. Your question is too broad to be answered appropriately within a paragraph IMO.

Good Luck.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

05-23-2016 02:35 PM

1) You can stick with the null hypothesis as long as you want. Statistics tell you about the likelihood that the evidence that you collected is in agreement with your null hypothesis. For example, if your null hypothesis is "I am going to win the lottery", statistics tell you that there is 1 in a million chance that this can happen. The choice is still yours to buy a ticket.

PG