Discovering Statistics

Information

This is a group for users of my textbooks (and anyone else really) to discuss the books, post questions and so on.

Members: 133
Latest Activity: on Tuesday

Discussion Forum

Statistics Questions 67 Replies

Use this to post statistics questions

Started by Professor Andy Field. Last reply by MJ May 17.

Model II multiple regression 4 Replies

Dear Real Statisticians (and I do mean psychologists): I am a Canadian archaeologist (BSc in Chemistry) researching the past human use of European Atlantic shellfish. After two decades of practice I…Continue

Started by Greg Campbell. Last reply by Marcelo Kittlein May 11.

SPSS Questions 12 Replies

use this to post questions specifically about SPSS

Started by Professor Andy Field. Last reply by Kyriakos Antoniou May 8.

Unsure of the correct analysis....

Hi Andy,This forum is a brilliant idea! I wondered if you would be able to recommend an analysis... I'm using a database for comparing men and women who have convictions of child maltreatment. I have…Continue

Started by Hannah Stone Apr 22.

View All

Comment Wall

Add a Comment

You need to be a member of Discovering Statistics to add comments!

Comment by MJ on Tuesday: Hi everyone,

I am doing ROC Curve analysis to establish cut-off points for a scale. I calculated Youden's index to establish cut-off points, however to my despair, the Youden's index is equal for two of the scores:

Let's say my subjects could score between 10-20 on a scale and the Youden's index is the highest at 124.5 for scores 15 and 16 on this scale. I wonder what would be the best cut-off: 15 and 16?!
obviously sensitivity is higher on one score and the specificity for the other

Thank you for your help

Comment by Henry Steel on Tuesday: Dear Andy and All

I have a question about data transformations. Consider the following explanation written by Andy:

“As such, transforming the data won’t change the relationships between variables (the relative differences between people for a given variable stay the same), however, it does change the differences between different variables (because it changes the units of measurement). Therefore, even if you have only one variable that has a skewed distribution, you should still transform any other variable sin your data set if you’re going to compare differences between that variable and others that you intend to transform.”

My concern is the following: I understand Andy's argument about changing the units of measurement, but what will happen if you transform a variable that is normal? Will it become skewed? If so, what is the use then of transforming the skewed variable(s)? Maybe I am missing something. Looking forward to some advice.

Best wishes

Henry

Comment by Rafael Garcia on May 1, 2012 at 22:48: Mutasem,

There's a decent description of post hoc tests here: http://pages.uoregon.edu/stevensj/posthoc.pdf

In short, different post hoc tests scale the F-ratio differently. Depending on how the F-ratio is distributed, you may end up with n.s. contrasts (common in Scheffe's method).

Comment by Mutasem Akour on May 1, 2012 at 22:27: Hello

I wonder how is it possible to get a significant omnibus F in one-way ANOVA, and getting no significant differences in a post hoc test ?

Comment by Mohamad Azwan Suroto on January 30, 2012 at 7:27: yey I am part of this group...

Comment by Rebone Gcabo on December 5, 2011 at 9:01: Thank you so Much Oscar, will try it.

Comment by Oscar on December 5, 2011 at 8:50: if you plan on sticking to SPSS, AMOS is the extension for everything Structural Equation Modeling that you should be looking for..

Comment by Rebone Gcabo on December 5, 2011 at 8:36: I need help with conducting MIMIC ( multi indicator multi factor analysis on spss. is it possible or do i need another software

Thank you