Homoscedasticity (con't)

This is the supplementary comment on homoscedasticity, see the original post below... (just 2 posts down).

Homoscedasticity is tested by a scatter plot of residuals. Residuals are errors, ie. actual score in population - predicted score. The line in the middle of the scatter plot is the mean of Y, ie. the mean of predicted scores.

If you get a rectangular shape of residuals around the line, that means that the residuals are evenly distributed around the mean. This is a good thing. Because homoscedasticity means equal variance of residuals, ie. the variance between the mean of y (the line, or y=0, which means no error) and the errors is homogeneous. So, getting a rectangular shape, which your graph may look like this...
(note: dots are residuals)
You can see that the residuals that is furthest away from the line above and below, from left to right have almost equal distance from the line. This is homoscedasticity.

In contrast, if you get a pattern, such as a "blow horn", which look like this...


You can see how the residuals actually fan out. That is a bad sign! The residuals that are the variance on the left hand side of the graph is small while the variance on the right hand side is huge. The variance of the residuals across the graphs are not homogeneous! So, the blow horn shape definitely violates the assumption of homoscedasticity, because this is heteoskedasticity.

Also, just a reminder, there is rule of tumb for you the check this assumption as well. The assumption is violated when the highest variance is 10 times bigger than the lowest variance, ie. with a ratio of 10:1.

Hope this helps.

Creating dummy variables in SPSS

Here is the command for creating dummy variables in SPSS. (from Dr. Mari Clements' notes)

transform-->
Recode into different variables-->
click over the categorical variable you need to recode-->
create a new name for the recoded variable-->
click change-->
click old and new variables-->
type in the old value (eg. 2)-->
type in the new value (eg. 1)-->
click add-->
repeat as necessary to get all values-->
continue-->
OK

Homoscedasicity

I have a question on interpreting whether our residual plot meets this assumption. In class, we were told if there was a pattern, the homosce. assumption has been violated. However, from other readings, it seems that if the points create a rectangular "scatter", it has met this assumption. I'm confused. Could you please give some further details about how to interpret this plot appropriately!

Thanks!

APA Publication manual (first printing)

Hello regression friends... (sorry for the nerdy greetings)

I would like to let you know that APA is now offering free replacement of the 6th ed. APA publication manual (first printing, which has tons of errors) with its second printing. This offer will end on December 15. So, basically, what you need to do is to call up APA and tell them that you will be sending them your first printing manual, and they will send you back the second printing. For details, please go to the APA publication manual first priting FAQ. (Just in case the hyperlink is not working, here's the url: http://www.apastyle.org/manual/corrections-faqs.aspx) They have to receive your first printing manual by 12/15/09. So, act fast!

Also, I just checked the bookstore today, fyi: they are selling the first printing. The second printing got out in August, so, I guess you may be able to get them somewhere around.

Here's the section from the APA website regarding to the replacement:

I purchased a copy of the first printing of the sixth edition of the Publication Manual from a source other than APA, but I still want a replacement. What do I need to do?

Even though you purchased your first printing of the sixth edition of the Publication Manual from another source, you are still welcome to participate in the return/exchange program. In order to receive an exchange replacement manual, you must return your current copy of the sixth edition to APA no later than December 15, 2009.

We would like to make this transaction as cost neutral to you as possible. A process has been created to provide you with a UPS postage paid return label that you can use to return the book to us. Please call the APA Service Center at 800-374-2721 for more information about the paid return postage process.

If you are interested in an exchange replacement copy, please contact the APA Service Center at your convenience for further information. Service Center staff can be reached at 800-374-2721 from 9:00 am until 6:00 pm (Eastern Time) Monday–Friday. E-mail correspondence can be directed to the Order Department.

How to make a new post on this blog?

1. You have to become a "contributor" of this blog if you want to make a post here. So, respond to the invitation that Dr. Kim sent you earlier this quarter to become a contributor.

2. Once you accept the invitation and become a contributor, log in on this blog with your user name.

3. Once you logged in, you will find a button says "new post" right next to your email on the top right hand corner of the blog.

Dummy Variables

Someone posted/asked:

1. I'm confused about what I'm supposed to be doing with the Dummy Variables. I followed the slides and I got to the point where I'm recoding agegroup into a different variable. But it seems like I'm supposed to be separating the one group, agegroup, into 4 different age groups. So I started transforming 3 into 0, reference group, and so on but that isn't right.

2. Where do I put the c1, c2, c3?

Response:
You're right on transforming! So, you need to recode each age group in order to run the analyzes. It seems like you are lost in the transformation... Let's illustrate the transformation with an example.

Let's say you're running an analysis on how the size of animals (large, medium, and small) predicts their daily diet intake.

So, in your original data set, you have size of animal coded as 1 = small, 2= medium, 3 = large.
Since you can only compare 0 and 1 in categorical regression, unlike gender (male, female), you need to more than one contrast coded predictors. Lecture 12 slide 19 addresses the issue of having multiple levels in one independent variable (IV).
# contrast coding variable = # of levels of the categorical variable examined - 1

In this case, there are 3 levels for size of animal (IV), so, you need 3-1 = 2 contrast coded predictors (aka. independent variables), and I will name the 2 contrast coded predictors as "C1" and "C2".

Let's say I'm going to have "small-sized animal" as the reference group, ie. small = 0, for both C1 and C2.
Then in C1, I will have medium animals as 1, anything not "medium" will be 0.
In C2, I will have large animals as 1, anything not large will be 0.

I will do the following...



So, C1, C2, and C3 are the names for the new predictor variables that will appear on SPSS. You will create C1, C2, and C3 as new variables in SPSS when you're transforming/recoding. Hope this helps.

Common Mistakes on Article Reviews!!!

So now that you have turned in 4 article reviews and received grades back, we wanted to note some common mistakes that were made that you should be careful for in the next 3 article reviews.

1. Unless they are general acronyms that anyone would know that looks at it, you need to write out acronyms. Only list acronyms when you need to use them again in your article review (AR). Also, make sure you type out the entire term when the acronym first appears in your AR, just like shown here.

2. Make sure you inclue doi or a website that anyone (without passwords) can access. The doi can also be found on the article page after you click on the title from EBSCO. For any articles that do not have a doi (which should be RARE), do a public search on google or google scholar. See APA manual for further details on finding doi.

3. While on the subject of doi's...Look up the proper way to reference the doi! There should be no space between the colon and the first number or letter of the identifier. Also, "doi" should not be capitalized. This is clear in the 6th edition references.

4. In general, pay attention to how to reference proper APA style. You are in a psychology graduate program, so you will need to know how to use APA properly for most (if not all) the papers you write here! It will and should become second nature but until then, triple check that all aspects are correct!
a.) Make sure to FULLY reference every piece...That means the correct initials for the names (if it gives a middle name or initial, it should be referenced with that middle initial)...that means the full title...that means the full and proper journal name...etcetera
b.) Make sure to space the initials properly, there is one space between initials!
c.) The order of the names in the article (believe it or not) IS IMPORTANT. Make sure to keep the correct order of the author names
d.) Make sure to watch the rules for article title and source title capitalization and italicizing. For a journal article, the first word of the article title is the only thing capitalized unless there is a colon, in which the first word after the colon is capitalized as well. The only other exception to this is you would also capitalize a word if it were a proper noun...think of this also like if you would capitalize a word mid-sentence, then you would also capitalize that word in the article title.
For the journal name, all words are capitalized and in italics
e.) the journal volume number is also in italics
f.) you include the issue number (the number in parentheses, which is NOT italicized) ONLY IF the journal paginates by issue. This means that you should do a search for the journal name, if you see that each issue it goes back to page 1, you WOULD include the issue number. An easy way to see this is if you see your page numbers go up above 200-250, it most likely does not paginate by issue and you would NOT include the issue number. However, if your article's pages are lower (1-200) either way you should check to see if the journal paginates by issue!
g.) already mentioned, but worth mentioning again! doi is NOT capitalized in the reference...also, there is NO SPACE between the colon and the identifier.

5. It's worth saying it short and sweet again: REFERENCE CAREFULLY AND FULLY!

6. Express numbers in numerical form when numbers are greater than 10 (e.g., 100, or 11). Express number in words when numbers less than 10 (e.g., nine, or two). Always express numbers in words when it is the first word in the sentence (e.g., One-hundred percent of the class is amazing).

7. APA recommends using 12-point Times New Roman font

8. APA clearly states that all words in a heading 4 or more words should be capitalized (e.g., Population From Which Participants Were Drawn)

9. For population we clearly stated earlier (and it is clear in the example) that you need to include the following information: age, number in study, gender, and ethnicity. This is for both quantitative AND qualitative studies. If some or all the information is not clear or not given, state so IN THE POPULATION SECTION.

10. These are meant to be one full page single spaced. Try to keep it as close to one page as possible, if you go 1/4 page over or under, we may have to take points off.

11. This is an APA style assignment (as is this class...again...until it becomes second nature you need to make sure to do it carefully), therefore, throughout the entire paper you must keep with APA guidelines. This is a lesson in summarizing so you don't have to re-read an entire article, so really quotation should be left out. However, if you DO quote (which, again is better not to) it is important (IMPORTANT) that if you use quotations that you cite properly. Look this up again...but for a random example (Labriola, et al., 2009, p. 234).

12. The relevance section is really for how this article is relevant for YOUR research interests. This section is where you would explain why and how it relates to your interests, not how it is relevant for the population. Also...I hope you realize that it is a waste of your time to read something and then do this AR if you aren't going to use it in the future!

For now, this is all. I know that it is a lot...but most of it has to do with APA style. This is not the only class that it is necessary to have APA style (especially referencing and in-text citing) correct! It will benefit you to learn it and practice it for all of your papers!

Questions? Further Comments?

--Your TAs

Article Reviews

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

Also, I need to mention again our class has a late penalty. "If any of the assignments are turned late, ..." Please take a look at the fourth page of our syllabus for more.

Clarification on variable selection + question (HW4)

1) For output on question 3, 4, 5, do we need to include age group into independent variable?
Although the instruction states that only first session outcome score and number of sessinos needs to be added, I was thinking perhaps the age group was left out.

2) If a variable is significant (p=.02) but the R2 of the scatterplot is almost close to zero (R2=.006), should the variable be left out of the model since the R2 is weak?

One more question

Sorry guys! I totally forgot about this question. In our APA manual, it states that it's okay not to list every individual variable in the table's title, as long as they are listed in the table. Is this how you interpret it??? Or should my title include all of the ind. variables we use in our steps??

Thanksgiving week

There will be no review session on the Wednesday of thanksgiving week. Happy thanksgiving!

asterisks

Just wondering if the asterisks should be in superscript or regular font.

Thanks!

ΔR^2

If we have ΔR² in our table, does the R² value for Step 1 go in the table or the note? Could the R² value technically be seen as a change from the null? Or is that an improper understanding of it?

Thanks!

P values

Hi TAs,

In our p note (* p < .05. ** p < .01. *** p < .001), if we underline the "less than" signs to indicate "equal to or less than" and then place asterisks accordingly in our Table, will we get points taken off? (I would have underlined here, but don't know how to underline on this blog) - Thanks!

Title for Table

Maybe somebody already asked this, but I scanned the blog and cannot find the question so forgive me if this is a repeat question...

If our title for our table goes beyond one line do we double space the extra lines or single space our title to make it look tighter?

Also, For the different steps, isn't the change in R squared the same for each variable of each step. In other words, wouldn't we have the same number for the variables that fall under each step? If so, should we go ahead and plug the same number in anyway-redundant as it may be?

THank you so much.

More questions! Sorry!

1. If our R square change gives us a three digit number past the decimal point such as .004, if we rounded that we would be reported R2 = .00...that's not correct right? We would need to put all 3 decimal places in this instance?

2.If we end up only having either p < .001 and p < .05, do we just use two stars in our note or would we say *p < .05, ***p < .001???

sorry these are so technical, but we're losing points b/c of this stuff and the manual does not give specifics esp. on tables!

Table spacing

On page 141 of the APA manual, it states that table may be submitted either single- or double-spaced...so, do we HAVE to double space it or does it just need to be readable?

Table

If we choose to put the R2 value in an additional column in our table, is it R2 or delta R2??

decimals

in the tables, do we round our decimals to second decimal place? .024--> .02 ; .636-->.64

Reporting significance level

Someone from the class asked how to report p value if it is great than .05 or equal to .05.

According to the APA publication manual page 114, report the exact p value when it is not less than .001.

When the significance level is .05, you will report p = .05.

When the significance level is great than .05, eg. .07, you will report p = .07.

When you are indicating significance level of values in your table, you still use * p < .05, * p <.01, or *** p < .001. You are not going to label any values with an asterisk that has a significance level equal to or greater than .05. Statistics which have significance level that is equal to or greater than .05 is not considered as significant if the model is using an alpha = .05.

Hope this helps!

Question regarding results paragraph

Do we need to indent the first line of our results write up? I got marked on my last homework for not doing so, but others that did not do it didn't. Also, none of the examples given by Dr. Kim have the first line indented.

sample write up for HW3

The prediction of college students’ global self worth from school competence, intellectual ability, social acceptance, physical attractiveness, athletic competence, and job competence was explored using a hierarchical multiple regression. The B, SEB, and β for each predictor at each step of the equation are presented in Table 1. The hypothesis that college students’ global self worth would be significantly predicted by school competence and intellectual ability was supported, R² = .38, F(2, 368) = 113.07, p < .001. Both school competence and intellectual ability significantly predicted global self worth, with higher levels of school competence and intellectual ability associated with higher global self worth. The addition of social acceptance and physical attractiveness to the model significantly improved prediction of global self worth, ΔR² = .27, F(2, 366) = 143.62, p < .001, with higher levels of social acceptance and physical attractiveness predicting higher global self worth. The addition of athletic competence to the model, however, did not improve prediction of global self worth, ΔR² = .00, F(1, 365) = 0.08, p = .78. Levels of athletic competence were not significantly related to global self worth. Finally, the addition of job competence did significantly improve the prediction of global self worth, ΔR² = .03, F(1, 364) = 28.78, p < .001. Greater perceived job competence was associated with higher global self worth. The overall model with all predictors was significant, R² = .68, F(3, 364) = 128.01, p < .001.

HW 3

When we report R2 for each step, which table to we get the F statistic, the anova table or the model summary?

my understanding was that the anova table in third block explained the signifcance of the overall model, where as the model summary explains significance within each step. . . .correct?

In our example Dr. Kim gave in class, Confidence intervals are not reported. Do we need to report them in our homework

Christie

Question for HW3

Okay, so I have a few questions:

1. If the p value is larger than .05, in the writeup do we just put p > .05? Or do we just state that the model was not significant and not provide any data at all? Basically, what do we do when the model does not change R2 and the pvalue is way to large?

2. If the R2 change equals .002, how do we report that on our table? As .00 or 0 or do we not report it at all?

3. When we are reporting the results in the writeup, for the second and third step do we use change in R2 or just R2?

4. When we are summing up and saying "the overall model," do we look at the Anova table to write up our results (eg The overall model R2 was significantly different than zero, F (7, 887) = 134.56, p < .001) (I changed the numbers) and if so, where do we find F on the anova table

Hierarchical MR write-up question...

Hey!
I was wondering how to write up the results for the HMR. I was thinking that it would be written this way but please let me know if this is wrong:

• you start with R2 for step 1 and its significance
• then change in R2 for step 2 and its significance
• and then change in R2 for step 3 and its significance
• and, lastly, for the overall model it would be R2 for step 3 and its significance....???

Thanks!

For students who were at the Wednesday evening review session

Hi guys,

Thanks for coming to the review session tonight. I was just going through the notes and I realized that I made a mistake...

When we were talking about which F ratio and df you are reporting for each step of a hierarchical multiple regression, I told you to report the statistics from ANOVA table. Whoops, I was wrong. You should be reporting statistics under the "change statistics" columns from the "model summary" table. Sorry for teaching you the wrong stuff, I hope it wasn't too late for me to fix it. Please let me know if you need further clarification and elaboration.

I will post the answers for those unanswered questions on the blog once I get a response from Dr. Kim.

Thanks!

Is homework 3 due Nov. 12 or Nov. 17?

Question we asked in class re: the homework

So I asked a question in class and Dr. Kim asked that I would post the question and his response on the blog.

My question was, how do I interpret the data output? How can I tell if there was a significant change in R2?
Basically his response was that if the p value is significant(meaning it is small) you can say that there was a significant change in R2. if the model does not increase R2 by much and the p value is not significant (meaning it is too large), then simply state that the model when such and such variable was added did not significantly explain additional variance- or something to that effect

Hope it was helpful!

Need clarification for using SPSS (HW3)

Do I have to insert corresponding predicting IV's from each hypotheses to each block in one process? (e.g IV from hypotheses 1 -> block 1; IV from hypotheses 2-> block 2, IV from hypotheses 3-> block 3; run statistics)

Or do I run three different process for each predicting hypotheses then insert IV that needs to be provided into the following block?
(e.g. predicting IV from hypotheses 1 -> block 1; IV needed-> block 2; run statistics
predicting IV from hypotheses 2 -> block 1; IV needed -> block 2; run statistics
predicting IV from hypotheses 3 -> block 1; IV needed-> block 2; run statistics)

Homework #2 & Quiz #2

I would like to let you know you can pick up your copies at the front desk.

Seong-Hyeon (Sung) Kim

TA's -

Can we use a few quotes in our article reviews when discussing hypotheses, participants, or results?

I'm finding it really hard to paraphrase some of these sections.

Homework # 3

Hi TAs,

I'm looking at my schedule and see that Hmwk #3 is due on Monday (11/9). I think I missed it being passed out - is it available for pick up somewhere? Thank you!

Articles not listed in the faculty recommended readings

Please attache a hard copy of the article you are reviewing if that is not included in the Faculty Recommended Readings when you submit your article reviews. So, if you are reviewing an article that is included in the Faculty Recommended Readings, you don't need to attach a hard copy, but only submit your review.

Best of luck to you!

CI value

I was wondering where we get the "t(N-2)(1-alpha) value for the CI? On our slides, the example gives 1.97 for this value...so are we to use this every time we need a 95% CI...what is the value then if we want a 99% CI?

Thanks guys!

DOI

2 of my articles do not have a DOI. What exactly do I do then? Do i just cite it as if i read it directly in the journal? or do I have to provide a link to the article? in which case it would look like this at the end of my reference (based on the link supplied by ebscohost):

Retrieved from http://search.ebscohost.com.naomi.fuller.edu:2048/login.aspx?direct=true&db=psyh&AN=2004-16129-004&loginpage=Login.asp&site=ehost-live

that feels like it is too much information, but it links you to the article if you are logged in i believe

christie

t test meaning

What does t test mean in words? I understand that t test is used when we do not know the sigma of the population. I understand that it doesn't tell us about the variations. But what does it tell us if our output says t= 4 and is significant? what does that mean about our data set (using a sentence)?

do you understand what i am asking?

Thanks

Christie

Policy on tutoring

This is the policy on 1 on 1 tutoring with our TAs. You could meet with a TA no more than once for one hour or less for 1 on 1 tutoring. Anything beyond that would really be inappropriate in the context of the other demands placed on TA time.

You are also encouraged to make a good use of study groups, discussions in class and me.

clarification on article reviews

1) The example article shows that the word "from" in the fourth heading is not capitalized. Is this a typo? (i.e. should it be capitalized?) Dr. Kim seemed to think "from" was not capitalized in class. Just wanted to get everyone on the same page.

2) Is there supposed to be a period after the DOI number? The examples in the APA book don't seem to have it, but it wasn't explicit so I wanted to make sure.

Thanks!

DOI for article reviews

So, in class we were told if a doi number wasn't available, we were to cite the website we retrieved the article from. However, the 6th ed. APA manual says on page 192 after the third bullet point additional info. about using EBSCO, which is where I will probably get my articles. I'm confused as to what this all means. How do I cite the URL properly from an EBSCO or PsycINFO article??? On page 191, 4th bullet point it states to provide the home page URL of the journal, book or report publisher...I'm computer illiterate...what exactly would be the home page in the library's EBSCO, etc??

Katie,

you had a really good way and understandable way of explaining the Central Tendency Theorem. . . could you repeat it for the blog?

What does the Central Tendency Theorem mean?

christie