Article View: sci.stat.edu
Article #4Statistics student needs help with linear regression
From: Mr. Proper
Date: Sat, 07 Sep 2002 23:24
Date: Sat, 07 Sep 2002 23:24
16 lines
1774 bytes
1774 bytes
I have a two linear regression questions:
1. I need to construct a confidence interval for
theta = beta_2 + beta_3 In multiple regression with 7
variables. I can get an estimate for theta (by just
adding beta_2 estimate to beta_3 estimate), but what
about confidence interval ? I understand that it has something
to do with the "linear hypothesis" F test ... or not ?
2. In another regression model (with categorical variables represented by
a number of dummy variables) I have to construct an F-test for the
hypothesis H_0 : categorical variable does not matter. I know
how to do this. But I'm also asked to show that this is actually
a generalized likelihood ratio test - and I don't have a slightest idea how
to do this...
Technical Details
Message-ID:
<8010766.0209071524.44787042@posting.google.com>Path:
arc1-ca.pugleaf.net!archiver1.google.com!postnews1.google.com!not-for-mailShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: fa931,e1ec61b02bbef9d6,start X-Google-Attributes: gidfa931,public X-Google-Thread: 109d8a,e1ec61b02bbef9d6,start X-Google-Attributes: gid109d8a,public X-Google-Thread: f732f,e1ec61b02bbef9d6,start X-Google-Attributes: gidf732f,public X-Google-ArrivalTime: 2002-09-07 16:24:34 PST Path: arc1-ca.pugleaf.net!archiver1.google.com!postnews1.google.com!not-for-mail From: dizzy2@bk.ru (Mr. Proper) Newsgroups: sci.stat.edu,aus.stats.s,sci.math Subject: Statistics student needs help with linear regression Date: 7 Sep 2002 16:24:34 -0700 Organization: http://groups.google.com/ Lines: 16 Message-ID: <8010766.0209071524.44787042@posting.google.com> NNTP-Posting-Host: 192.91.75.29 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: posting.google.com 1031441074 30222 127.0.0.1 (7 Sep 2002 23:24:34 GMT) X-Complaints-To: groups-abuse@google.com NNTP-Posting-Date: 7 Sep 2002 23:24:34 GMT