Chi square value

Hi all,

I am trying to perform gaussian fitting to a raman peak.....However, i am doing this with and without weighting wave to see how it changes. I know that in the case of no weights, the chi square value that igor gives is just the sum of the squares of the residuals. I purposefully introduced a random weighting wave (standard deviation for each point on the data) and i got a value of 1.1. Now i have the following questions: 1). Is that now the reduced chi square value or do i have to divide by the degrees of freedom? I am asking this because at the moment the value indicates a very good fitting.....If i divide though it will give me something way less than unity which is probably not good. Do i need to divide or this value (1.1) is already the reduced chi square? 2). The standard deviation of each point when we introduce weights is probably related to the y axis (intensity) values because the x axis points are stable (same increments during Raman acquistion). Do you have any idea of how i can calculate errors for each data point between the fitted and the experimental values? I am still confused of how to come up with errors for each data point. I hope it makes sense. Thanks

Konstantinos