HW grades??

Hello,

I know this might be a long shot, but I was wondering if there was anyway the TAs can grade the factorial and repeated measures hw prior to the exam?? I know it's a lot of work, but I believe there are a few of us who would like to know if we got the hw assignments right or not so we better prepare for the exam. If it's impossible to have all the hw assignments done before the exam, can we have an answer sheet to the hw so we can see what we did right/wrong? Thanks in advance for considering this plea. :)

DO NOT create a new turnitin account

Someone created a whole new turnitin account for the Repeated Measures assignment (rather than changing the pseudonym on his or her existing account, as instructed). This results in there being two accounts for that person on turnitin.com, delinking everything done prior to this assignment from everything that will be done from that assignment forward. Note also that this does not protect your confidentiality, because your old pseudonym comes up with no completed assignment. If you were the person who did this, please log back in to turnitin.com under your original account, click on the "user info" tab at the top, scroll about 2/3 of the way down, and change your pseudonym. You can also change your email account on this screen, should you so desire. Once you have done this, upload your Repeated Measures homework and printout to your original turnitin.com account (the originality statistics will show 100% overlap, but this is completely understandable given the circumstances). Thanks.

ANOVA Assumptions

Let me know if you need the handout of assumptions from today's review session and I'll email it to you.
Gemma
gemmapartlow@fuller.edu

Hey I was wondering how much does the purple in Turnitin have an impact on our grade? to be more specific when the articles are graded as well as the HW, there is purple highlights, are these accounted as part of what we get for our grade?

Grade Not Posted

I have not received a grade for either my article review 3, or my factorial ANOVA hw. I know that other students got their grades for the article review 3 a while ago, so I just wanted to see if there was a reason mine is not posted. I was hoping I would know how I did before the next article reviews are due, since I use the feedback to make improvements.

Thank you very much!

Exam 2 Review Sessions: Treats, but no Tricks :)

Exam 2 is approaching, and review sessions have been scheduled. In order to better ensure that all who would like to attend will be able to do so, there are two review sessions scheduled. Students may attend either, both, or neither, as you see fit. The first one is on Monday, 10/31 from 5pm-6pm in Room 311 of the Psychology Building. The second one is Tuesday 1pm-2pm in room 130 of the Psychology Building.

Are the limitations we are supposed to discuss about ways that the study itself was limited or how the analyses on SPSS were limited?

In the HW since sphericity was assumed since the variable only has 2 level and only 1 pair can be created, in the write up do we say that sphericity was assumed and then indicate the Tests of within subjects effects between the factor and the greenhouse geiser, however that wouldn't make sense especially because the test of within subjects reveals significance. How do we report this. What kind of stats do we put in our write up since our mauchy's test of sphericity shows no significance.

Blog Posting

I keep getting emails and comments on the blog from folks who are confused why they can comment but not post. The reason is that the blog WAS set up to allow comments without being logged in, but new posts could only come from registered, logged in members of the blog (and nine of you still haven't accepted your invitation to join the blog). As this situation was clearly confusing to people, I have changed the permissions on the blog so that you will no longer be allowed to comment without being logged in. If you have lost your login information, and need me to resend your invitation to the blog, send me an email from the address you would like to log on with. If this doesn't cover your situation, and you are positive that you have both accepted your invitation to the blog *and* are currently logged in from the email address that is registered with the blog, and you still can't post, please contact blogger help.

repeated measures and assumptions

ok so I have a few questions about the repeated measures and mixed design.

For fully repeated measures homogeneity of variance is assumed if sphericity is assumed, correct? But how do you test for normality? the q-q plot and box plot didn't show outliers but the histogram didn't look completely normal. I know that we can use the equation to test skewness and kurtosis but I thought that that test was only used in small samples? is n= 38 small?

For a mixed design- do you test for normality the same way as fully repeated measures mentioned above (just looking at time and not the intervention group)?? How do you test for homogeneity of variance? I'm assuming time (within) is checked with Mauchly's test, so does that mean that for intervention group I can just look at the Levene's test that the output produced after running a mixed design for repeated measures? If so, can I just assume that if Levene's test is not significant than homogeneity of variance is assumed?

When, on SPSS, I go to Analyze-->GLM-->Repeated Measures, I feel like I am doing something incorrectly.  When I put label the "Within Subject Factor Name" and the levels for the time and intervention and click "Define," there are not enough options in the left side box for me to define the terms in the "Within Subjects Variables" on the right.  I feel like I am missing something.

On turnitin.com, my second article review no longer says that the grade is pending and will let me view the page where it usually shows comments, but it is blank where there should be a grade.

Is this a glitch or just because it hasn't been graded yet?

assumptions for Repeated Measures ANOVA

For the repeated measures ANOVA, do we only have the assumption of sphericity? Or do we also have the assumptions of normality and homogeneity of variance? I noticed in an earlier post that we test normality in the same way, but do we need to do this?

homogen. of var. for repeated measures

I received output that 'no homogeneity of variance b/c no between subjects factors' when running SPSS on repeated measures. Does this mean that there is no homog. of variance assumed in repeated measures, or am I doing something wrong?

homework

how do we run a 'mixed model ANOVA'? Does this mean we need to create new variables to see the interaction effects, or run a 'multivariate' analysis under general linear model or something else?

Thanks!

Resources Added to the Blog

The data set for the in-class demonstrations for Repeated Measures has been added to the blog.

The Repeated Measures homework (clarifying the submission process is through turnitin.com) has been updated.

Have a great week!

normality in RM

How do you examine the assumption of normality in repeated measures designs? I assume this is not the same as sphericity (?)

Let's say hypothetically that p is actually equal to 0.001, and not p < 0.001.

What do we report?

Do we round to 3 decimals and put p = 0.001?

Or leave it at the standard 2 and put p = 0.00?

Exam 1 at the front desk

The first exam is at the front desk for your review. (Remember, you cannot take it with you, but will need to review it in the waiting area.)

I encourage you to look at the questions you missed and make sure that you understand why you missed them, as each new analysis you learn in ANOVA assumes that you understand the ones before.

The grades on the first exam were quite good, with a mean of 88...

In the factorial ANOVA writeup example it is written (under the discussion of orthogonality of predictors), "If cell sizes are equal, it can be understood that the variables are not correlated to each other." What is the logic behind that?

For questions 1 and 2 on the homework...do we just look at the QQ plots and box plots to determine normality? Or do we need to do some type of mathematical computation as well?

Ques. #4 in HW

Should we write our answer in APA style using the results of the post-hoc's or should we just simply answer the question AND THEN report the results in the write-up?

Updated Article Review Suggestions

The revised Faculty Suggested Reading List is posted on the class moodle site. Please do check it out, as many faculty members updated their suggestions!

In truthers

Can the term "outlier" ever be defined as a point that is not similar in type to other data points? (Like a real number in a data set of whole numbers, for example). Or does "outlier" always mean a point that does not fall within the normal distribution of points?

Descriptive Statistics

In addition to a table with the descriptive statistics incorporating each level with each IV, should we report in the write up itself the descriptive statistics for the main effects found? That is, the table represents the statistics broken into the categories and groups within themselves, but should we report them as a whole within the write up (e.g. males (M... SD... n...)... females (M... SD... n...))?

Factorial Homework Instructions

Many thanks to the students who pointed out typos in the factorial homework.

A revised pdf has been posted on moodle. If you've already completed the homework, don't worry, the corrections do not impact the analyses. They simply clarify that SPSS printouts should be submitted to turnitin.com (thus preserving your anonymity) and point you to the right question in answering #8.

Please do review the revised factorial homework directions before submitting your assignment.

post-hocs

does a separate one-way anova need to be run (after a factorial anova has been run and based on significant differences found) in addition to running a post-hoc to examine where the differences lie?

Factorial ANOVA Hw

I keep getting no significant interaction for either Lifequal or Lifequal2. I don't think this is correct, since we are asked to run post hoc tests, but I am not sure what I am doing wrong. I did the practice homework correctly.

I also noticed that I get a warning when I open up the data set that says something like the data set is written in a character encoding (Windows- 1252). I'm not sure if this matters or not.

Should I only be getting 5 points on my Q-Q plot diagram?

Main effect without significant post-hocs?

In a factorial ANOVA is it possible to get a main effect for a particular factor but when post-hocs are run no significant pairwise differences shows up? If it is, what do you do then in terms of reporting it?

Negative Skew/Kurtosis Values

what do I do if I get negastive values for skew or kurtosis, when the errors for either/both are positive? Dividing it out to get the z will always result in something less than 3.8 or 3.3 or 2.5 (e.g. 10 divided by -0.5 = -20, a large number, but still less than the significance values). Is it an absolute values thing? Did I do something wrong? Do I even need to worry about this?

Factorial anova

I ran the factorial anova and the post hoc comparisonsfor the three variables listed on the assignment sheet. I received an error message for run a post hoc on the gender variable. What do I do with this information?

Oneway Homeworks Graded

All oneway homeworks have been graded and are available on turnitin.com.

You will want to check the rubric for which points you received and which you lost. You will also want to carefully read the feedback in the comments.

Should you have questions after doing these things, make an appointment to meet with me by emailing my assistant, Stefene Lister at stefenelister@fuller.edu.

If you turned in your assignment late, please note that the penalty for lateness does *not* show up on turnitin.com. You can calculate your final grade for any late assignments by applying the penalties laid out in the syllabus.

When should we expect to get our exam results back?

My major finding in the article was so extensive, I have attempted to include mostly everything but will I lose points if I don't?

Article Review p value question

In some articles we review the authors may list the p values as p < .05 or p < .01 instead of giving exact p values. In our papers though if we reported them this way it might appear as a APA style mistake when actually it is from the article itself and we can't be more specific. What should we do about this?

Article Review doi

If my article does not show a doi, is it okay to simply not list it? Is there something I can do to find the doi?

Pseudonym Use

As noted in your syllabus and as discussed in class, EVERY assignment you complete for this class is submitted under a pseudonym. Putting your real name on an assignment defeats the purpose of that system.

I realize that putting your name on your paper is a habit, but it is one that you need to break for this class.

Failure to follow this direction on any and all assignments due after today will be penalized.

Article Reivew

In the Article Review, Article 2, the results section does not indicate significance only percentages and graphs. So what shall I say in regards to my result section about significance. The kind of test used in my study was Multiple Logistics Regression.

Equal Variances Not Assumed

When reporting a t-test, do we need to inform the reader if equal variances are not assumed? Or do we just leave that out since the test controlled for that?

When reporting the degrees of freedom for a t-test where equal variance is not assumed, how do we round it? For instance, my output says df= 20.873.

Do I write t(20.837) = 82

or t(20.84) = 82

or t(20.8) = 82

?

Thank you!

clarification about factorial ANOVA

Hi, everyone, I just wanted to clarify that on the review sheet we passed out today, I double-checked with Dr. Clements, and you will not be asked on Exam 1 about factorial ANOVA (part of # 7 and also the significant interaction effect on #10) because it has not been covered yet.

On another note, I still have a few copies of the review session hand out, so if you find me (Charlotte) tomorrow, I can give you a copy. Happy studying!

Sum of Squares

I understand how to compute the sum of squares, but what does that value tell us conceptually?

Question from Review Session: Power

Okay, so there was some confusion about power during the review session, so hopefully this clarifies. I'm looking at pages 18-19 in the book, so it would probably be helpful to go read that before asking further questions. There's a handy chart there, and power is in the lower right section of the chart.

If you reject the null hypothesis, when the null hypothesis is in reality true, you make a type 1 error. The symbol for this is the greek letter alpha, and alpha is the probability of making a type 1 error. Often alpha is set at .05, but it can be adjusted.

If you accept the null hypothesis, when the null hypothesis is in reality false, you make a type 2 error. The symbol for this is the greek letter beta, and beta is the probability of making a type 2 error.

Since ANOVA is interested in whether there is in fact a difference between groups, and this is represented by correctly rejecting the null hypothesis (which is that there is no difference), we call this power. Power is represented by the expression (1 - B), or 1 minus Beta.

So, yes, absolutely, Power = 1 - Beta. My apologies for any suggestion otherwise.

You can read up on strategies to increase power, and their strengths and weaknesses on page 19. Also consult the slides to round that out as well.

decimal places and zeros reporting

When reporting for t-test and CIs, do we report 2 decimals always, even if one number behind the decimal is a zero? Ex. t(102) = 5.20, p < .001, with 95% CI [12.70, 28.36]. Should this instead be t(102) = 5.2, p < .001, with 95% CI [12.7, 28.36] ?

article review

I know that we can review one non-empirical study this quarter, but can it be a case-study, where the methodology is unstructured review of the client's responses to socratic questioning from the therapist to evaluate the client's improvements?

output

i realized that i ran the wrong post-hoc test for anova, for the output would it be okay to have two post-hoc tests displayed even if i am only reporting on one, or should i delete it?

Reporting CI

I found an example for reporting confidence intervals:

89% CIs [3.45, 2.7], [-6.0, 3.89], and [-7.23, 1.89]

Is it okay if we group the confidence intervals for each t-test in brackets like this? Also where and how should we report this in the text?

The notes say that Homogeneity of Variance is only sensible to assume if each group is providing an estimate of the same population variance. What would be an example of when Homogeneity is not sensible to assume because each group is providing an estimate of different population variances?

effect size reporting for t-test

Does the effect size need to be written in our report for the t-test. If so, how do we write that up (ex. with 95% confidence that the true population mean lies between 0.01 to 10.01)?

fmax

does an fmax need to be computed for a t test or anova only when levenes is violated? thanks!

APA format

I am having a hard time finding information on how to format the results section in APA format. I've looked throughout the APA Publication Manual and haven't found a place where it really explains everything. Please let me know if there is a place where it explains both the rules for significant figures and the APA format in general for results.

Thanks!

Decimal points

Just to clarify:

Do we just round to the next decimal point past what was measured?

How come there would be a significance difference between two levels but that difference does not exist when a post hoc is ran?

In regards to the Write Ups, do we only do 1 writeup for the independent sample t-test or do we do 3 write ups in response to the difference between the groups?

SPSS Results

Would we lose HW points if we include histrograms and Q-Q plots?

Regression Approach on Test?

Will we be tested over the anything having to do with Regression as it was covered in the book, and in the unedited notes?

One-way Anova write-up

should the t tests and ANOVA result write-ups be in an individual paper? and should the rerun oneway ANOVA questions also be in a seperate paper as well as the answers to what our variables represent?

test question

Will we have to memorize formulas for the test (e.g. to find Fmax, z score, Sum squares, Mean Squares, DF, Partial eta Squared, Standard deviation, Variance, etc) OR will they be listed and provided for us on the exam?

What is the difference between complex and pairwise types of comparisons?

Homework

Do you calculate an Fmax for a t-test? If so, how would you do this?

Outliers

What is the easiest way of checking for outliers?

Q&A for Homework

Is it possible to have a Q&A or some type of review session (where we can ask questions, receive help on the homework) that will help in doing the homework correctly? I feel I am not the only struggling to be sure that I am doing the homework correctly. Thanks.

Test

What will be the format of the test?

Collaboration

Are we allowed to collaborate on the practice problems?

Posting to the Blog

Note that you must be a registered author to create a new post to the blog. That requires that you accept the invitation to the blog that you were sent on 9/27. Twenty-three of you have not yet done that, and so are not going to be able to create new posts. I have re-sent invitations to all 23 unregistered folks. Accept them, sign in to the blog under that registration, and you will be able to post.

Again, this restriction is in place to prevent the blog from being spammed.

Question Regarding Homework

I did the practice homework correctly, but for some reason I am having a really hard time starting the actual homework. I am trying to "select cases" and do the t-test between two levels but every time I select cases and put in LT 2 or LT 1 it crosses out all of them. How can I select the right levels?

Thanks!

Obtaining variances

As a reminder, to obtain variances by group (for the purpose of checking Fmax, for instance), you will want to follow these steps:

Choose "Data" from the menu across the top of SPSS.
From the Data drop-down menu, choose "Split File."
Click "Organize output by groups," and click over the name of your Grouping (or Independent or Between Subjects or Between Groups) variable.
Click "OK."
This will return you to the main screen, and sort your data into groups, which will each be analyzed separately.

Then choose "Analyze" from the menu across the top of SPSS.
From the Analyze drop-down menu, choose "Descriptive Statistics" and then "Frequencies."
Click over the name of your Outcome (or Response or Dependent) variable.
Uncheck the "Display frequency tables" box.
Click on "Statistics" and then choose "Variance." [While you're here, note that this is also where you would request "skew" or "kurtosis."]
Click "Continue."
Click "OK."

This will send you to an output screen, where your variances for the dependent variable will be listed by each level of the independent variable. From there, it is a simple step to pick out the largest variance and the smallest variance, and calculate Fmax.

Once you have done this, restore your data set to its original, unsplit condition by
Choose "Data" from the menu across the top of SPSS.
From the Data drop-down menu, choose "Split File."
Click "Analyze all cases, do not create groups."
Click "OK."
This will return you to the main screen.

Requests for Additional Help

If you think you currently need additional help with the class, please submit your request for that help to me, not to the TAs. Send me an email at clements@fuller.edu.

This is the only way that I can make sure that all students are given equal and fair access to additional help. Should you need help beyond the time available to you from the TAs, further assistance from the departmental teaching fellow is available, but again, you will need to request that directly from me.

Review Session for Exam 1

The TAs will be holding a review session for the first exam on Tuesday, October 11, from 1-2 p.m. in Room 314 of the Psychology Building.

Another question about the book - From Chapter 2

This may be silly because it's from Chapter 2, but...
Pg. 61: Where is the data for this page?
Pg. 63: Why did we change the value from 21 to 1 in this example if it was a common typographical data entry error?

**when calculating error in an ANOVA, is it represented by the within groups or between groups part of the equation?

Question about the book

Where can I find a supplemental reference guide for ANOVA definitions and common equations?

Pseudonym Use

Read all of this post, even if you think you used a pseudonym correctly.

I am very serious about the use of a pseudonym. Both the TAs and I want to evaluate each assignment you submit as objectively as possible, and being unable to connect the work to previous or future ensures that we can do so.

Unless some of you thought that your classmates' names would make great pseudonyms, several of you did not submit your first assignment under a pseudonym. Let me be very clear that although the assignment was accepted this time, future assignments that are not submitted under a pseudonym will not be accepted and will receive failing grades. That is how serious I am about the integrity of the grading process and the protection of your identity.

Consider that the TA evaluating your work in this class might be your classmate in another class, or your fellow committee member in PGU, or your lab mate in your research lab. Your work in this class is not relevant to those settings, and you deserve to be able to work with your fellow grad students without any concerns that they have access to your grades.

If you made a mistake in signing up for turnitin.com and entered your real name in the registration process, you should correct this mistake by logging in to turnitin.com, going to the "User Info" tab at the upper right of the screen, scrolling halfway down the page to the "Last Name" and "First Name" fields, and putting in your new pseudonym. Note that it should NOT be the one you used in the in-class, baseline assessment, because I already have access to that key.