help interpreting regression results with dummy variables

[largeselection]

So I ran a regression on different "type" scenarios and got the following. I see that the R-Square is not a great fit (<95%) but I was hoping I might be able to use this regression to figure out some relationships and make some determinations about which "Type" would yield the best results. In the results the coefficient for "Type F" is this highest, so would I interpret that to mean that if an observation is of Type F then it could be expected to have the highest results? The only weird thing is that when I pivot the original data on type and results, "Type F" does not have the highest Sum of the Types, nor does it have the highest average. So why would the regression put that type as having the strongest positive relationship?

Sorry if I'm being confusing, but I don't have so much experience with regressing in excel (and almost no experience using dummy variables).

Thanks for any help, links, etc to help me understand this better.

Excel Workbook
	A	B	C	D	E	F	G	H	I
3	Regression Statistics
4	Multiple R	0.883330456
5	R Square	0.780272694
6	Adjusted R Square	0.776373937
7	Standard Error	179.6441043
8	Observations	659
9
10	ANOVA
11		df	SS	MS	F	Significance F
12	Regression	8	74605235.27	9325654.408	288.9704138	2.4365E-208
13	Residual	651	21009074.73	32272.0042
14	Total	659	95614310
15
16		Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
17	Intercept	0	#N/A	#N/A	#N/A	#N/A	#N/A	#N/A	#N/A
18	Type A	216.8795181	19.71850217	10.99878258	6.3098E-26	178.1599782	255.5990579	178.1599782	255.5990579
19	Type B	370.1847826	18.72919257	19.76512235	1.80984E-68	333.4078656	406.9616996	333.4078656	406.9616996
20	Type C	363.5666667	18.93615126	19.19960723	1.94085E-65	326.3833625	400.7499708	326.3833625	400.7499708
21	Type D	316.2241379	23.58842652	13.40590216	2.30919E-36	269.9055579	362.542718	269.9055579	362.542718
22	Type E	308.9204545	19.15012586	16.1315104	1.68984E-49	271.3169867	346.5239223	271.3169867	346.5239223
23	Type F	439.8080808	18.05491181	24.3594699	1.16961E-93	404.3551914	475.2609702	404.3551914	475.2609702
24	Type G	275.2978723	18.52887459	14.85777622	3.26397E-43	238.9143026	311.681442	238.9143026	311.681442
25	Type H	322.6545455	24.22320608	13.32005947	5.70907E-36	275.089503	370.2195879	275.089503	370.2195879
Sheet15

 

[sapka]

I believe that is not the right forum to ask these type of questions.
As far as I remember statistics, you equation is meaningless. All the coefficients are insignificant according to p-values, and the equation itself is also insignificant according to F-stat.

 

[largeselection]

Ok, I thought I had seen some other questions similar to this on this forum at some point since it was regarding the regression function in excel. I will look for a statistics forum to post in as well. I also believe the opposite is true of what you posted regarding P-values and F-statistics. Since lower p-values indicates a lower chance of obtaining that result (which would be the null hypothesis). So we would reject the null (which is that there is no statistical significance between the two) and accept that there is below 0.05 or some such. This is the same case for the F-statistics I believe.

Haha, again I'm trying to remember statistics like you are too.

Thanks for your input.

 

[sapka]

You are right! I have just checked, the lower the p-value and F-stat value the better.