Help: how to do the lack of fit test using SPSS software?

[wxmwqr]

Hi, everyone:
Recently, I submit a parper about the quantitative method using HPLC.
The one of reviewer ask me to do the Lack of Fit test on calibration curve.
I have the statistical software "SPSS", but I don't know how to use it to do the statistical analysis. Is anybody can help me?

An other question about statistical analysis.

When I read a paper, the author used the statistical method to evaluat the between-day precision. But I can not understand why and how to do that.

I quote the the article, and hope some expert give me reply.

"One-way analysis of variance (ANOVA) was carried out with grouping variable "day" to assess precision at 95% level. The result was non-significant when data for each day was compared with other days and within the same day."

[bartjoosen]

Hi,

I have no experience with SPSS, mostly R, cause it's free.
(www.r-project.org)

But I did a Google search:

http://instruct1.cit.cornell.edu/Course ... 19lab3.htm

"For SPSS users, go to Analyze-->General Linear Model--> Univariate... Then designate your outcome variable as the DV and your predictor variable as a COVARIATE. Under the Options click the Lack of fit test."

Good luck

Bart

[wxmwqr]

thanks, bartjoosen
But it is realy hard problem for me.

[bartjoosen]

An other question about statistical analysis.

When I read a paper, the author used the statistical method to evaluat the between-day precision. But I can not understand why and how to do that.

I quote the the article, and hope some expert give me reply.

"One-way analysis of variance (ANOVA) was carried out with grouping variable "day" to assess precision at 95% level. The result was non-significant when data for each day was compared with other days and within the same day."

You can split up your variation (precision) in 2 different variations: day to day variability, and the variability within a day. Then you can compare the variation within a day with the variation between days and see if it's significant at a 95% level. I'm sure SPSS can do the trick for you, but as stated before, I'm not an SPSS user.

For more statistical help, you can look at the nist e-handbook:
http://www.itl.nist.gov/div898/handbook/index.htm (home)
http://www.itl.nist.gov/div898/handbook ... mpc114.htm (variability)


Good luck

Bart