article review ?

Ok...for one of my article reviews i have run into a problem...

When searching the results section (and then the rest of the paper!) i could not find the models used specifically stated. I know that they found frequencies, but the rest of their results are simply given in different symbols. I can only speculate (and try to remember) what these symbols mean. A few of the statistics look like this: x^2(l) = 8 54, p < 01 ....which i BELIEVE is chi-squared? Then further down it changes to Z = 3 25, p < 001...which i BELIEVE is z-test?

Are these speculations correct? should i simply state that they didn't give a specific model used in determining results?

If needed, here is the reference for my article:

Egeland, B., & Sroufe, L. A. (1981). Attachment and early maltreatment. Child Development, 52, 44-52.

Thanks in advance for your answers!

Review Session Questions...Answered!

Hi all!

As promised, I contacted Dr. Clements to get the answers to questions I wasn't sure about that were asked during tonight's review session.

1. Sobel's Test is reported with the symbol z prime. That is, use a lowercase, italicized z with a tick mark superscripted above it. (If I could get the blog to show it, I would.) If you don't understand from my description, you can see the article that Dr. Clements suggested you look at for an example of writing up mediated models found in an earlier post on the blog.

2. Always, always, always use 2-tailed tests of significance.

3. You do NOT have to include ANY figures for the moderated/mediated models homework. However, IF you do choose to include them, make sure they are formatted according to APA format.

4. Finally, we come to centering. This is important for one of the methods for testing moderated models. If you choose to test it using the method that requires an interaction term, you need to center the continuous predictor variable. Make sure you do this before creating the interaction term. Otherwise, you'll have collinearity issues.

SPSS GradPack Installation Issues

Bought the gradpack this weekend. I had a lot of problems trying to get the software to work. After waiting 1 1/2 hours for SPSS tech support this morning, it turned out that the company mislabelled their installation authorization codes.

In the gradpack, you will see two authorization codes: SPSS Installation Authorization Code and Amos Installation Authorization Code. For my gradpack, the code for the SPSS Installation Authorization Code is actually the code for AMOS Installation Authorization Code, and vise versa. If you are thinking of purchasing the SPSS software, and you encounter similar installation issues, you might want to try using the Amos code.

The SPSS technician said this problem has been noted and that I wasn't the first person to have this problem. Check with the SPSS technician if you are still having installation issues.

Good luck.

Logistic Regression Yanked

I have decided that it makes more sense to consolidate what we already know about regression than to add another topic...

Sobel's Test

Regarding to question 2. A., how do we report Sobel's test result in APA format? Is there a specific symbol stand for Sobel's test, such as B stands for unstandardized B? Thank you.

Sobel's Test, Part 2

I think I understand what the diagram on the website means conceptually, but my question is more related to reading the SPSS printout. On my SPSS output for the third regression of question 2, I have three Betas: the constant, and two Betas for the two independent variables I used in the regression. What I'm not sure of is which of these three Betas Sobel's test is asking for.

Sobel's Test

I'm having trouble figuring out which numbers to use for the online Sobel's Test calculator (http://www.danielsoper.com/statcalc/calc31.aspx). On the website it asks for four values: A, SE of A, B, and SE of B. I figured out where to find A and SE of A (i.e. the Beta and SE of the predictor-mediator regression), but I'm not sure where to find B and the SE of B (i.e. the Beta and SE of the mediator-outcome regression). According to the "Additive, Moderated, and Mediated Models" handout, slide 11, there is no Mediator-Outcome regression, so where do I find the Beta and SE that the Sobel Test is looking for?

Centering in SPSS

i cannot seem to find directions for the keystrokes necessary to center a variable in SPSS. How does one translate the equation into something that SPSS recognizes- i've hit a wall. I've done it manually but am not sure whether or not the results are precise enough.

2 questions

(1) When looking at slide 7 and realizing which test we should complete, i realize that the one we use does not include looking at the interaction. I'm wondering, then, if we still need to do so because in slide 9 the second bullet says that we must interpret the interaction. So, do we still need to look at the interaction term even though it is not discussed in the type of test we are using for question number 1?

(2) In the hw you say that "if" we include figures...this doesn't say that we need to, but should we since it is mentioned in the instructions?

Thanks!

Betas for interaction model

After adding the interaction term to the second model, it seems unnesscessary to explain the betas for the gender and PAQ variables because they have already been explained through the interaction term interpretation. Is it then necessary to interpret the Betas for gender and PAQ for the second model in the write up after you have already explained the interaction term?

At Long Last

Received from the bookstore
----------------------------------------------------------
Dr. Clements,

The SPSS 17 Grad Pack is now available in the bookstore. We currently have 5 copies for Mac and 3 copies for Windows. Thanks!

Jeremy Mast
Bookstore Floor Supervisor

Based on the last question:

Can we use a calculated interaction term for question 1 with gender, by treating gender as continuous?

Furthermore, can we center gender if we have problems with collinearity? Is it okay to center gender?

Multiple vs. Linear Regressions

I'm looking at question #1 on the homework and I'm not sure exactly what type(s) of regression(s) to run. According to the PowerPoint slide handout for "Additive, Moderated, and Mediated Models", Slide 7, we are supposed to run three separate regressions:



1. Outcome by Predictor

2. Outcome by Moderator

3. Outcome by Moderator/Predictor Interaction Term



Did I understand this right? Also, are we running three separate linear regressions, or one multiple regression, or some combination of linear and multiple regressions?

Sample moderation/mediation APA write-up

Looking through the notes, I notice that we do not have a sample for how to write up a moderation or mediation results section like we had for Multiple Regression/Linear Regression. From the notes, I know what moderation and mediation mean, I just don't know how to write them up. I know how to use APA format, I just do not understand how to say what I need to say and what statistics are appropriate for moderation/mediation.

Hierarchical Stepwise Regression

What is a hierarchical stepwise regression? Is it as just as "bad and evil" as stepwise regressions? How does it relate to forward and backward entry?

Thank you.

Formula for the Curve

For any assignment curved in any of my classes, the formula is the same:

Individual's score
------------------------- X 100
Highest score in class

In this case, there were a few folks who scored 100 or 92, but I took them off the curve, and used the next highest score (90) as the denominator.

So, if your score had been 80, solving this equation, you would have a score of

80
--- X 100 = 88.8
90

Next Quarter ANOVA Sections

The ANOVA class is just like the Regression class in terms of the course sections. The two sections are the same--there is no PhD or PsyD section. Choose the one that fits best with your own work, class, and personal schedule.

Q regarding the article review

Hi,
I read an article using the quantitative reserch review method, and wrote a review under the empirical article review format. I just wonder whether it's ok, or I'd better adopt the theoretical review format.
Thank you.

Recommending normality transformations...

So, the book doesn't recommend the need for transformation if there is a problem with normality (p.46). I think they give a reason for this, but if it's their reason, i don't understand it. They even say that "many regression textbooks" do find this type important. I know transforming is largely subjective...but what is the rule of thumb for this?

Thanks in advance for the clarification!!!

Interactive and Mediated Models Homework

The homework assignment is posted on Portico. Note due date change from syllabus.

Article Reviews and Next Homework Assignment

So, given that this coming Tuesday has many, many things due, I thought that it might make sense to modify our due dates as follows:

The Article Reviews that were due on Tuesday are now due on Thursday.

The Interactive Models homework that was due on Thursday is now due on the following Tuesday.

I realize that this just redistributes pain rather than eliminating it, but if it helps you to have a couple of extra days, please take them as outlined above.

df1 = k - 1

Hi, I'm working on question 9 in the review sheet.

When using the equation df1 = k - 1 in regression, is the constant considered one of the predictors so that df1 = k - 1 = 2 (the constant and another predictor) - 1 = 1?

Is this the right way to look at it or should I be taking the class again?

?? regarding the text

Okay...So, i've been meaning to ask for a while, just never sat down to ask.

The book makes a lot of weird marks above the symbols, and i dont' know what the different marks mean. For example...The predicted value from the model is (symbols that i can't even try to make with word so i'll attempt to explain in words)...X_{i with a tilda (~) above it and then} β with a ^ above it...

p.48 for reference due to my lack of vocabulary...

So...do these symbols have a meaning? I'm sure they do...so i guess what is the meaning they carry? and is it common to use them?

Review Sessions this Week

In addition to our regularly scheduled Monday session at 5 p.m. in room 126 Weyerhaeuser, there is an additional session scheduled for Wednesday at 6 p.m. in room 311 Weyerhaeuser.

(Weyerhaeuser is the official name of the Psych Building...)

Study Guide for Exam 2

The review sheet for Exam 2 is posted on Portico.

Sample size & descriptive statistics

I was about to write up my results section for the homework and noticed that the homework handout mentions surveying 240 college students, while my SPSS output looks like it has a sample size of 371, where F(2, 368) = xxx.xx for model 1. Was this a typo on the handout, or am I missing something?

Also, do we need to include descriptive statistics in our results write up? Thanks!

write up

For question 1A, we are asked to report whether adding more predictors will improve the predictability of all of the X variables or not. Do we have to report the R square for each model or do we have to report the R square for the overall model only? When the change of R square between two models is 0.00, does it mean adding extra predictor variable(s) does not improve the prediction? Thanks!

question about the write-up

As I am looking at my powerpoint notes, I noticed that the write-up talks about "higher education weakly but significantly predicted lower happiness" and later "Age was also found to signicficantly predict happiness, with older respondents slightly but significantly less happy." I am assuming these are referring to the unstandardized betas and obviously the significance values.
My question is how else to state/explain the differences in unstd. betas as we look at six different variables. I started to say in my own write up that one had a slightly stronger correlation than the other or that the strongest variable correlated was such and such. I'm afriad this is too close to the dreaded "better predicts" that Dr. Clements told us to avoid and that correlated isnt the right description??
Is it important that we differentiate between unstd. betas as we do our write up, or is it enough to say that they all significantly predicted self worth or all but such and such sig. predicted?

Thanks...

Transformations PowerPoint Slides

Hi all!

After looking back over the PowerPoint slides from the transformations lecture today, I realized that I made a change to a slide, but forgot to mention it. Your print-out of the slides needs to be modified at one point.

If you look at the first slide on centering, it states "To center: Y i centered = Yi – Mi." It should actually read, "To center: X i centered = Xi – Mi" We only ever center predictors, and X stands for our predictors, while Y stands for our outcome variables. To center, subtract each value of the predictor from the mean of the predictor (or some other meaningful value). Sorry for the confusion.

What means altering the order of entry of the blocks?

For the second step of the HW, we have to "alter the order of the entry for the blocks of variables". Note it does not say altering the order of variables.I am interpreting "altering the order of the blocks" as we should keep together the variables that were together in a block in the first analysis. In other words, if in block 1 of the first regression I had variables A and B together, in the second regression I will keep A and B together but will entered in step (block) 2,3,4. Is this correct? Should I mix the varibles and create new blocks (for example have varible A in step 3 and B in step 4).

Colinearity and Tolerance

When describing colinearity for part C of question 2, we are using tolerance statistics via SPSS (which have an inverse relationship with with colinearity; e.g., the higher the tolerance value, the less colinear are the variables). SPSS gives tolerance statistics for each variable in every step (or block) of the model. My question: for each specified variable that is excluded (in the first step there are 4, the second, there are 3, and the third, there is 1), does the tolerance stastic associated with that variable refer to colinearity between that variable and the model at that point in time? For example, at Step 1, the tolerance level that is reported for physical appearance is the relationship of colinearity between that variable and the overall model- which up to that point has used intellectual ability and school competence (and therefore only refers to colinearity between these variables used together and physical appearance?) If you could help to deliniate what the tolerance levels mean for each block, it would be helpful.

Chris

Difference between the first hypothesis and 2-4.

In the write up portion of the assignment should we account for the difference in wording from the first hypothesis and the other three. The first states "expect to predict higher levels of global self-worth", while the other three state "significantly improve prediction of global self-worth". Does "improve" refer to the "prediction" or "global self-worth"?

The way I read it is that "improve" is referring to prediction, so the variables in hypothesis 2, 3, and 4 could all improve prediction for lower self-worth and still prove the hypothesis. Am I reading this wrong?

Multiple Regression Homework: Question 1

Are we running four hierarchical multiple regressions (i.e., 1 - block 1; 2 - block 1 and 2; 3 - block 1, 2, & 3; and 4 - block 1, 2, 3, & 4)? Or, are we running one hierarchical multiple regression (i.e., 1 - block 1, 2, 3, & 4)? The coefficients table yields slightly different values.

Thanks.

hypothesis 1

For the first hypothesis of the Multiple Regression homework, are we using a simultaneous multiple regression to find our answer instead of a heirarchical?

Or are hypothesizing that perceived school competence and intellecutal ability would predict higher global self-worth, holding all other variables constant? So if it were a heirarchical regression, would the independent variables in the first block 0f the heirarchical regression be schcomp and intabil, and the second block all the other variables that we are considering for the assignment?

Writing Workshop

Please encourage any and all students who are having difficulty with their writing skills to attend our upcoming Writing Workshops, presented by Fuller's Writing Center. By encouraging your students to attend, you will not only help them to improve the quality of their writing, but also potentially save yourself and your TAs some grief in grading papers at the end of the quarter! :) All workshops are taught by Fuller alumna Kathryn Streeter, who came to Fuller with six years of English teaching experience and has, to date, served as a TA for a total of nine classes taught by four different Fuller professors. The workshop schedule is as follows:

Composition Workshop
How to develop a strong thesis statement, form coherent paragraphs, and use supporting evidence appropriately
Friday, November 7, 2008, from 1-3 p.m., in Payton 102

Advanced Composition Workshop
How to develop different kinds of thesis statements, transition deftly between paragraphs, smoothly integrate quoted material, and more!
Friday, November 14, 2008, from 1-3 p.m., in Payton 102

Grammar & Editing Workshop
How to fix common grammatical errors such as inconsistent verb tense, flawed sentence structure, and misplaced punctuation
Friday, November 21, 2008, from 1-3 p.m., in Payton 102

All workshops are FREE, and no RSVP is needed. A packet of helpful materials will be available at each workshop for $5.

Thank you for your help in getting the word out! Flyers will be available in classrooms this coming week for the benefit of interested students. Feel free to contact kathryn_streeter@fuller.edu with any questions.

Trouble with downloading SPSS file

I was trying to download the SPSS file for our Multiple Regression assignment in the library computer lab and it did not work. I went back to the linear regression assignment SPSS file, only the zip file would work. So, can we can have the zip file for Multiple Regression as well? Thanks!