SSR + SSE = SST. The coefficient of determination = SSR/SST
- Thread starter fashepard
- Start date
Hi Frank,
It seems to be working fine at the moment...however:
1. If it was a peak load issue, I am thrilled to say we are currently tranferring the media files to secured, state of the art cloud facility at @ nirvanix. As this host is scalable, we should *never* have peak load issues. I currently have good hosts but now the heavy media will be hosted at, based on my research, the world's state of the art cloud facility. I've spoken to some nirvanix customers with massive media files, way bigger than ours (e.g., feature length movies) and they've had fantastic experiences....currently testing (today), should deploy in 1, 2 or 3 days max....no impact on premium page...so, soon the site host plus media host will represent the best possible technical solution we can find (i.e., i have shaved no pennies to deliver the media).
2. Re SSR+SSE = SST, you are not using Gujarati's terms. (although you are correct in substance).
Please note, I published the early birds *before* GARP released the Study Guide.
Rather than go to early bird, you should review this under the regular schedule (Quant A - C) b/c the regular schedule indeed reflects Gujarati's terminology, which is the assigned.
(This is also why any candidate should be careful, nay reluctant, about re-using 2007 material)
Pleaes note (Spiegal was assigned stat last year)
Explained variation = Spiegal's SSR = Gujarati's ESS
Unexplained variation = Spiegal SSE = Gujarati's RSS
Total variation = Spiegal's SST = Gujarati's TSS
So, there is no point in memorizing Spiegal's terms (they aren't in Gujarati Ch 7...), you care about:
RSS + ESS = TSS
and R^2 (coeff of deter) = ESS/TSS = 1 - RSS/TSS
Okay, but those are the terms and we want to be "current" about them, but of course, the substance is the same and the graphical understanding of the relationship is important too
David
It seems to be working fine at the moment...however:
1. If it was a peak load issue, I am thrilled to say we are currently tranferring the media files to secured, state of the art cloud facility at @ nirvanix. As this host is scalable, we should *never* have peak load issues. I currently have good hosts but now the heavy media will be hosted at, based on my research, the world's state of the art cloud facility. I've spoken to some nirvanix customers with massive media files, way bigger than ours (e.g., feature length movies) and they've had fantastic experiences....currently testing (today), should deploy in 1, 2 or 3 days max....no impact on premium page...so, soon the site host plus media host will represent the best possible technical solution we can find (i.e., i have shaved no pennies to deliver the media).
2. Re SSR+SSE = SST, you are not using Gujarati's terms. (although you are correct in substance).
Please note, I published the early birds *before* GARP released the Study Guide.
Rather than go to early bird, you should review this under the regular schedule (Quant A - C) b/c the regular schedule indeed reflects Gujarati's terminology, which is the assigned.
(This is also why any candidate should be careful, nay reluctant, about re-using 2007 material)
Pleaes note (Spiegal was assigned stat last year)
Explained variation = Spiegal's SSR = Gujarati's ESS
Unexplained variation = Spiegal SSE = Gujarati's RSS
Total variation = Spiegal's SST = Gujarati's TSS
So, there is no point in memorizing Spiegal's terms (they aren't in Gujarati Ch 7...), you care about:
RSS + ESS = TSS
and R^2 (coeff of deter) = ESS/TSS = 1 - RSS/TSS
Okay, but those are the terms and we want to be "current" about them, but of course, the substance is the same and the graphical understanding of the relationship is important too
David
