Notice: On March 31, it was **announced** that Statalist is moving from an email list to a **forum**. The old list will shut down on April 23, and its replacement, **statalist.org** is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Christopher F Baum <baum@bc.edu> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
re:st: RE: RE: Insignificant coefficient in prediction |

Date |
Wed, 1 Dec 2010 16:55:50 -0500 |

<> Thanks. I share with you the skepticism on significance. Let me ask this way. Let's say a practitioner wants to find which factor has more impact on sales to make an investment decision: store size or good location. Let's say, a regression model tells that the effect of store size is smaller than the effect of location but highly significant and that the effect of location is larger but insignificant with p>0.5, for example. What should we recommend? To invest in store size or to pick up a better location? I guess he should invest in store size. Then, what implication does it have on prediction using both coefficients? Are these two problems very different ones and should not be mixed? If your data cannot pin down the coefficient on location -- or even, perhaps, its sign -- then you probably should respecify the model, dropping that variable, and predict from the new model. But just because the data cannot give you a precise estimate of the effect of location does not mean it is irrelevant -- and there are probably other studies that have found it important. You need to worry about how you have measured location; it may be that some alternative measurement (or a measure of location taking other factors into account) will give you quite different results. For instance consider selling takeout food on the northbound side of a divided highway leading out of Boston (such as US Rte 1 north of the city). You have to drive 3-4 miles out of your way to get to a store just across the road. Commuters pick up food on their way home after work. A location on the east side of the highway (with the flow of homebound traffic) might be expected to do much better than the ! same store just across the road, where it will not attract much of the homebound commuting traffic. So how do you measure that 'good location' vs 'bad location' factor? On the other hand a Dunkin Donuts or Crispy Creme on the inbound side might be much more successful. Nick and others, keep in mind that in the colonies we (usually)* drive on the right side of the road, and don't have those pesky roundabouts every mile on major highways. Kit * This is a general rule, and the general rule in Massachusetts regarding driving is that people don't pay that much attention to general rules. Kit Baum | Boston College Economics and DIW Berlin | http://ideas.repec.org/e/pba1.html An Introduction to Stata Programming | http://www.stata-press.com/books/isp.html An Introduction to Modern Econometrics Using Stata | http://www.stata-press.com/books/imeus.html * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**Follow-Ups**:**RE: st: RE: RE: Insignificant coefficient in prediction***From:*"Jeff" <jbw-appraiser@earthlink.net>

- Prev by Date:
**st: RE: RE: Insignificant coefficient in prediction** - Next by Date:
**st: Change in coefficient sign with Pooled Fractional Probit** - Previous by thread:
**Re: st: saving residuals for many regressions** - Next by thread:
**RE: st: RE: RE: Insignificant coefficient in prediction** - Index(es):