Help! Stats background required for Grad School

[clinpsychjack]

Today was my first day of graduate school in a Clinical Psychology Ph.D. program. We were told to read Ch. 1 of my Stats book before class and, while I got through it, I'm worried that I don't have adequate background to do well in Statistics.

I asked my professor for some supplemental resources to help navigate the text, and he said "This is a fairly advanced course with students, who have quite a lot of background." He said that if the course becomes too "painful" I should take undergraduate stats courses offered within the Dept. The major problem is that this would completely derail the course progress I need to have by 2nd year, and I don't feel I would be prepared (in terms of statistical knowledge) to propose and develop a Master's thesis, which is due by the end of the Fall semester of my second year.

If you're in a Psych grad program and you've gotten through stats, what kind of statistics courses did you taken before grad school?

Just some background: I've taken Intro to Probability, Statistical Methods in Psychology, and a Calculus in college...and have never been a strong math person. I did complete a Honor's thesis in Psychology, but I was able to navigate that with SPSS. The Stats course I'm currently enrolled in uses SAS, which I (and many others in my class) know nothing about.

Any advice?

---

Members don't see this ad.

 

[erg923]

I really just had a basic backgrounds in stats (1 college level stats class). I got a masters first (before my doctoral program) and remember being scared to death about my stats class. However, for some reason, I do not remember it being all that difficult. And we did go through it all, ANOVAs, MANOVAs, ANCOVA, MANCOVA, contrast coding, regressions, etc. My doctoral level stats classes included few computations actually (only on the first test I think) and were very big on understanding the underlying theory. At the doctoral level, long, drawn-out, step by step computations are probably not going to be a part of the class. Lots of reading and quantitave reasoning is really what the focus will be on. Coming from a person who is also bad at math, I found the focus on the underlying theory (i.e.,the quantitive reasoning part) to be much easier to handle than the jumble of formulas and numbers and computations i was forced to do undergrasd level stats classes.

 

[myelin]

I took undergraduate courses in basic statistics and statistics for the social sciences. I too was overwhelmed on the first day of statistics in the MA program I'm in. I hadn't had the material in over 3 years! It truly was a "welcome to grad school" moment.

Everything turned out to be ok in the end. I spent extra time at the library, buried amongst piles of SPSS output. See if you can't get some help from another graduate student or a GA/TA. Most professors schedule the first test before the drop date. I'd see how you do and go from there.

 

[Ollie123]

First...how did the CLASS go? How are the homeworks? Does the prof have notes/powerpoints?

Many stats books are just plain terrible (*cough* HAYS *cough*). I've done very well in my stats courses so far, though I'm obviously still chugging along and it will be awhile before I'm comfortable with plenty of aspects of it. I still can't understand a word in one of my textbooks. Reading it actually seemed to hinder my understanding of the class, so I stopped using it. Occasionally I'll look up a table or something basic, other than that it just collects dust on a shelf. I'd have freaked out too during my first stats course if I thought the book was an accurate reflection of the material, but it wasn't even close.

I've never been happy with a stats course, even though I typically do fine in them. My learning style for that material just always conflicts with how people like to teach it. I want the concepts first, then the "utility" (i.e. interpretation), then the math at the end to help me connect the two. I've yet to find someone who did that for me, so I largely end up self-teaching.

 

[Therapist4Chnge]

I want the concepts first, then the "utility" (i.e. interpretation), then the math at the end to help me connect the two.

I had 2 undergrad psych classes, so by the time I took stats in grad school I felt okay, but still concerned because it was a number of years in between my classes and grad school. I just spoke with my stats professor last week about this exact progression. He taught concepts first, showed the problem, and then walked us through the interpretation. It took me longer because I need to first get the concepts down, understand the HOW/WHEN/WHY, and *then* do the calculations. In the end I felt more grounded in the material because my focus was on the concepts and reasoning....and the math is just the math.

 

[deleted176373]

Today was my first day of graduate school in a Clinical Psychology Ph.D. program. We were told to read Ch. 1 of my Stats book before class and, while I got through it, I'm worried that I don't have adequate background to do well in Statistics.

Just some background: I've taken Intro to Probability, Statistical Methods in Psychology, and a Calculus in college...and have never been a strong math person. I did complete a Honor's thesis in Psychology, but I was able to navigate that with SPSS. The Stats course I'm currently enrolled in uses SAS, which I (and many others in my class) know nothing about.

Any advice?

Suck it up, I thought I was good at stats when I rolled in... only to have my butt handed to me. You'll need to enjoy the steep learning curve, find a student in your program who has a handle on this and get help quick! Remember you're supposed to struggle, it's grad school.

Mark

 

[JockNerd]

I tend to agree with Ollie; sometimes stats classes are just taught poorly and books are written badly. The courses you listed should be sufficient, I would think, for a first year grad psych course.

SAS is a bull at first, that might be contributing.

 

[Ollie123]

I had 2 undergrad psych classes, so by the time I took stats in grad school I felt okay, but still concerned because it was a number of years in between my classes and grad school. I just spoke with my stats professor last week about this exact progression. He taught concepts first, showed the problem, and then walked us through the interpretation. It took me longer because I need to first get the concepts down, understand the HOW/WHEN/WHY, and *then* do the calculations. In the end I felt more grounded in the material because my focus was on the concepts and reasoning....and the math is just the math.

And yet I'm constantly amazed at the number of stats profs out there who teach you how to plug things into formulas without ever telling you why, what it means, what assumptions need to be met, what to do if those assumptions aren't met, etc. A psych prof in my undergrad did exactly this, and I've met many people who had similar experiences with undergrad stats. My business stats prof was slightly better, but not by much.

At the end of my first undergrad stats class, I could do an ANCOVA by hand, but I still had no idea what an ANCOVA was, why you would do one, or what any of the numbers I got meant. I was able to pick that up on my own eventually but still...what a waste of a semester.

Glad to hear I'm not the only person who learns this way though. Math doesn't bother me in the slightest, but I need more.

 

[Therapist4Chnge]

I ended up using a couple of different books, a couple of websites, and a binder of notes that has been passed down from friends of mine. Of course....the binder went missing a year later, and the website links died with my last HD formatting, so I'm only left with the stinkin' textbooks again!!

IIRC stats textbooks were talked about recently in a thread. I remembering asking if anyone had a good suggestion for an SPSS book.

 

[myelin]

And yet I'm constantly amazed at the number of stats profs out there who teach you how to plug things into formulas without ever telling you why, what it means, what assumptions need to be met, what to do if those assumptions aren't met, etc. A psych prof in my undergrad did exactly this, and I've met many people who had similar experiences with undergrad stats. My business stats prof was slightly better, but not by much.

At the end of my first undergrad stats class, I could do an ANCOVA by hand, but I still had no idea what an ANCOVA was, why you would do one, or what any of the numbers I got meant. I was able to pick that up on my own eventually but still...what a waste of a semester.

Glad to hear I'm not the only person who learns this way though. Math doesn't bother me in the slightest, but I need more.

Depending on the number of variables, that could get really nasty, really quick.

 

[Cosmo75]

SAS? Ew. I'm sure it's a lovely program if you take the time to learn it. I have to admit I haven't looked at its interface in several years, so I'm hoping they've made it a bit more user friendly? Though based on JockNerd's comment, doesn't sound that way.

We had 2 stats courses, though you could proficiency exam out of Stats I. The 2nd was multivariate, regressions, etc. I had an excellent prof and book, so I feel like I was pretty fortunate. Going in, I only had 2 undergrad stats courses (one in the psych dept, one in the math dept). Mind you there had been a 7 year gap between grad and undergrad, so I was a bit rusty. Like others have said, it's a steep curve but most people are in your boat. You'll pick it up.

 

[Boston2k]

I don't think it's unnatural to not understand all of the material at first in stats. It's not like reading a novel, where you can immediately put the concepts together. Take good notes, make good friends (they probably have the same problem you do, especially if the prof is using SAS and they've never used it), find the TA who will probably have more patience/time than the prof, and get notes from previous years' students. It can take a couple of years applying the stats to various projects to get it down - don't feel like you have to understand everything completely at first, but make sure you see everything (i.e., read the book, listen carefully in lecture, take thorough notes). As you review it a 2nd-3rd-4th time it will make more sense, and it probably won't click completely until you start applying it to real data sets. From my personal experience along with those who have been in classes with me, your experience is not uncommon.