Thursday, December 10, 2009

Table Note: Double or single spacing

I agree with Jay. In the APA manual (p. 141), "tables may be submitted either single- or double-spaced." I am not able to find about a specific requirement on table note, though.

I recommended double-spacing. However, single-space won't get a penalty.

Wednesday, December 9, 2009

double spacing note

in the APA manual, it says that the note under the tables SHOULD NOT be double spaced-ie, the note should be single spaced. However, many agree that Dr. Kim said that the everything should be double-spaced and we were not marked off for it on our homeworks.

Should we double space the note or single space the note?

Rebecca Kulzer

Homoscedasticity and the Good Enough Rule

Hello - last minute question!

Regarding assumption 5 and homoscedasticity, I think I understand that the rule of thumb is that the assumption is violated when the highest variance is 10 times bigger than the lowest variance, ie. with a ratio of 10:1. However, on the "Assumptions in Regression" handout, the Heteroscedasticity Example does not have a variance 10 times bigger than the lowest varience, but there is heteroscedasticity and assumption 5 is violated. Could you clarify? Thank you!

Tuesday, December 8, 2009

Question on titles...

On homework 3, I missed a point b/c I left out the various specific variables in the title as they were included in the table...however, I've been told that it is redundant to say something like "Speed Predicted by Anxiety, Time Constraint, etc" in our time IF anxiety, time constraint, etc. are the variables listed in the table on the left most column...could someone please clarify so I know which to do?

Thanks!!

Monday, December 7, 2009

Another question

On our table #2, if we included a sentence or two in our write-up about the transformation, can we also put something like Height = centered height...etc. in our note for the table...or do we not need to???

Also, in my table I've listed Height^2 as one of the variables...is this ok or should I say "height squared"?

another question

on our second table, do we need significance (aka ***) for each value-B, SEB and beta-or just for the beta???

Answers to your questions

One of the TAs summarized and forwarded your questions to me. TAs, thank you so much! These are my answers.

1. When reporting desriptive statistics like beta, standard error, etc... in our APA table when the number is 0.001 (for example) would we insert 0.00 or would we include 0.001?

0.00 not 0.001

2. Also, although it doesn't need to be APA, for multicollinearity statistics do you want us to give the specific numbers or just whether the numbers meet the cutoffs?

Copy and paste the table that lists muliticollinearity statistics and state whether they meet the cutoffs.

3. Could someone provide an example of a title for question 1...how do we state correlations between, SD and M without being redundant of info provided in the table?

descriptive statistics (or means, standard deviations) and correlations of study variables (please do not copy and paste this. you need to format it according to the APA styles.)

4. For question 3, do we need to state why we "decided" to use the square method to transform...we were not taught how to decide which to do. Can we just say something to the effect of "the plots show we need to transform____?

Yes. Please read the question carefully. The answer is there. (hint: scatterplot)

5.Do we need to report the exact values for VIF and tolerance when we talk/write about multicollinearity?

Please refer to my answer to Question 3 above.

6. In our write-up for number 6, how do we report transformation? Could we please have an example of that...also, do we need to report anything about cheight, sqheight, or sqCheight anywhere in that write up?

"In the analysis the height variable was centered. Also, a quadratic term of height was created by squaring the centered height variable, which, then, was entered into the regression model."

7.Do we need to paste the bivariate partial plots for number 8? aka the cheight or the sqcheight?

I haven't asked for any partial plots for number 8. Please read on the blog my answers about the assumption checking for Questions 2, 5 and 8.

8.So my understanding is that we only round to 2 decimal places on tables. What do we do with a number like 0.00049? SPSS by default shows 3 decimal places. So that would be shown as 0.005. Rounding to 2 places would make that 0.01. However, looking at the original number, we would get 0.00. These seem like quite different numbers. Do we put a note somewhere in the table?

Please refer to my answer to the number 1 above. Also, look at pp. 113-114 APA manual.

Quiz 3 at the front desk

Your quiz 3s are ready to be picked up at the SOP front desk.

Attach "all printouts"

Because the total SPSS output could be very long, you don't need to attach everything. However, you need to copy and paste things as follows:

Questions 2, 5 and 8: relevant plots (assumptions)
Question 3: scatterplots
Qustions 4, 5 and 8: table that shows multicollinearity

Saturday, December 5, 2009

More questions

1. do we need to report R in our write up for number 6? Or just R2 or delta R2?
2. How do we get SPSS to show us the exact result for p...sig?
If we find SEB, etc. values that are .000 or .0002, how do we report them in the table??
3. the * for p values...does it matter if we just hit shift and ** or do we need to make them a superscript...does it matter???
4.I'm lost on what to report in my note for the table for question 7. Do we need to say that height means centered height? Do we need to report the delta R2 if we didn't report them in the table but reported them in our write-up? I was told this was redundancy on my homework #3?

rounding on tables

So my understanding is that we only round to 2 decimal places on tables. What do we do with a number like 0.00049?

SPSS by default shows 3 decimal places. So that would be shown as 0.005. Rounding to 2 places would make that 0.01. However, looking at the original number, we would get 0.00.

These seem like quite different numbers. Do we put a note somewhere in the table?

outliers

according to our notes, the cut off is determined by 2k/n for large sample sizes... what is k?

Question about final project

1. Could someone provide an example of a title for question 1...how do we state correlations between, SD and M without being redundant of info provided in the table?
2.For question 3, do we need to state why we "decided" to use the square method to transform...we were not taught how to decide which to do. Can we just say something to the effect of "the plots show we need to transform____?
3.Do we need to report the exact values for VIF and tolerance when we talk/write about multicollinearity?
4. In our write-up for number 6, how do we report transformation? Could we please have an example of that...also, do we need to report anything about cheight, sqheight, or sqCheight anywhere in that write up?
5.Do we need to paste the bivariate partial plots for number 8? aka the cheight or the sqcheight?

Friday, December 4, 2009

No Reviews During Finals Week

Hey All!

I just wanted to let you know that there will not be any review sessions during finals week. However, if you have questions feel free to ask here on the blog or email and we will respond on the blog!

--Your TAs

Reporting Descriptive Statistics

When reporting desriptive statistics like beta, standard error, etc... in our APA table when the number is 0.001 (for example) would we insert 0.00 or would we include 0.001?

Also, although it doesn't need to be APA, for multicollinearity statistics do you want us to give the specific numbers or just whether the numbers meet the cutoffs?

When significance is .001 and we do not know what the numbers mean when we click on them, what should we report p < .001 or p < .01?

Thank you so much!

normality of error plot not required for Question 5

One of the Mon-Wed students told me that in that class I told the students that only two plots for homoscedasticity and correct form of relationship between DV-IVs are required for Question 5. So, I am not going to require a normality of error plot for Question 5. I recommend including a normality of error plot for Question 5, but if you don't, you won't be penalized for that. No worries!

If you have any other questions, please let me know.

Testing Assumption 3 (Correct form of relationship between DV-IVs

Analyze-Regression-Linear, then specify your model. Then, click on the Plots button. Next, move Dependent into Y and zpred into X. In that way, you can test the correct form relationship. You need this plot for Questions 2, 5 and 8.

APA styles graded

You need to provide APA-formatted tables for Questions 1 and 7 and an APA-formatted write-up for for Question 6. For these qustions, APA formats will be graded. For the other questions, you don't need to follow APA styles.

my office hours

Hi Class,

I know you have some questions regarding the final project. Please let me know those questions. I will do my best to answer those questions. However, if you think you need my help, please visit me during my office hours. I am going to provide extended office hours next week.

9 am - 11 am, Monday Dec. 4, 2009
9 am - 11 am, Tuesday Dec. 5, 2009
9 am - 11 am, Wednesday Dec. 6, 2009

You can stop by my office during those hours, bu I strongly recommend to schedule a session with me by e-mail before your visiting.

Sung

testing assumptions

For assumption checking (Question 8), I want you to test homoscedasticity, the DV-IV relationship and normality of errors using relevant plots. "Furthermore," evaluate multicollinearity for these questions. For Questions 2 and 5, test only homoscedasticity and the DV-IV relationship.

Thursday, December 3, 2009

Final Project Question

sorry to have so many questions... I just wanted to confirm something... in class he told us that we only need to use APA format for the tables and for the write-up in number 6. -- are we going to have to use APA format for any of our other answers, even if we are talking about measures of, for example, multicollinearity? I want to make sure that I don't lose points over something I can prevent...
Quick question: If you are entering something in a note on an APA table, and it is describing the variables or just noting something that was done to the data (like centering it) does that have to be italicized or have any specific format?

Wednesday, December 2, 2009

APA Style Resources

Hi everyone,

I showed a website on APA style at review session today and promised that I will blog it... here it is....

http://www.docstyles.com/apacrib.htm

Also, you can check the Purdue Owl website too! http://owl.english.purdue.edu/owl/resource/560/01/

This websites have been updated after the 6th ed. APA publication came out. Hope these will be helpful resources for you. :)

Plots to test assumptions

I'm a little confused on the which assumptions are being tested & with what plots? Can you correct me/give me more info on the following:
Assumptions 1 & 2 we don't look at any plots for
Assumption 3: we look at a partial regression plot...is it the same one as for homoscedasicity??
Assumption 4:I have no idea!
Assumption 5: Homoscedasicity...a residual plot...is this the same as a partial regression plot?
Assumption 6: Histogram and Normal p-p plot?

Also, what is a bivariate scatterplot & is this different from a partial regression plot? do we ever use a bivariate scatterplot?

HW4 ready to be picked up

Your HW4s are ready to be picked up at the front desk.

Question about Cook's D chart on project

The Cook's D on question 4 (with transformed variable) look almost identical or perhaps is identical to the one in question 8 (centered and transformed variable).
Is this suppose to happen?

Tuesday, December 1, 2009

Regression with a Categorical Independent Variable (UPDATED!!!)

Hi everyone,

Since I noticed that some of you will have a quiz tomorrow without getting your HW4 back, I would like to provide some feedback which may help your quiz tomorrow.

If you can still remember that you were asked to calculate group mean for different levels (age group) of an independent variable, some of you might have rounded off the unstandardized betas before plugging them into the formula that help you figure out the group mean. I suggest you not to round off unstandardized betas before you plug them into the formula in order to get a more accurate group means. For example, you can see that you got an unstandardized beta (4.5678) on the coefficient table (spss output), plug in 4.5678 instead of 4.57 into the equation to get a more accurate group mean.

Hope this helps!

All the best with the quiz tomorrow or Thursday!! :)

UPDATE-->
I would like to add something too.

For the HW4 question 3 asking you about whether both first session and number of session significantly predict DV or not. You need to look at the significance level in the coefficient table (spss output), but not assessing the significance level by looking at R square only. The significance level is determined by whether the beta of the IV reaches statistical significance or not.

I know this is late at night... I wish I could have put this up earlier.

I'm praying for you guys...

APA stuff

If the Journal title has an ampersand (&) used instead of "and", should I type out "and" in my reference or should I keep the & symbol?

Also, is PTSD known widely enough that I can leave it at that or should I spell it out?? It's part of the title of the journal article.

HW3 ready to be picked up

You can pick up your HW3 at the front desk.