Week of Aug 24

After the Week of Frustration last week, things are beginning to pick up. We have some data from Hot or Not, and are testing here and there.

Had a huge breakthrough in the Value Learning E-Prime programming today (thanks to my neighbour who knows Visual Basic) and now have a working version of a program that will present Gain and Loss stimuli, add five cents for a correct response, take five cents away for an incorrect response, and update the total after every trial. Victory! Still working on a version that will handle Neutral stimuli - hopefully that will be ready soon.

Spent a whole day yesterday with Ben working on the SD Review. We revised the outline to make more sense and flow better, and I am much happier with it now. There are a few sections where Ben has papers and just needs to add the info to the wiki, so am just waiting for him to do that and then I think we will have most of a draft. I feel like we are much further forward after yesterday.

Started brainstorming ideas for NSERC applications. I think I will end up pitching the Value Learning stuff. Need to work on this more next week.

Emailed Jan at Stonehenge twice, but no reply. I guess we will leave that for now.

Things I learned this week:

1. Tabbing matters when programming Inlines. Things that are tabbed in will run only in the current Procedure (local). Things that are flush to the left side of the window will attempt to run globally - ie. they will run through once when the experiment starts and not every trial.

2. Declaring everything as an integer = BAD IDEA if you are working with decimals.

From Kamil:


There are a series of different types:
Bit, or boolean (1 and 0, or true and false)
- Dim x As Boolean
Integer (+ve and -ve whole numbers)
- Dim x As Integer
Unsigned Integer (+ve whole numbers)
- Dim x As UInteger
Floating point values (decimal numbers)
- Dim x As Single
Double values (decimal numbers, higher precision, higher range)
- Dim x As Double
ETC...

Once again, floating point values are sufficient for basic calculations.

http://msdn.microsoft.com/en-us/library/47zceaw7(VS.80).aspx
Gives a detailed description of what exists, and what are the specifics of each.

Mark & Anne Meeting Aug 20: Week of Frustration version

- Went into long-winded explanation of recent E-Prime problems with script not generating

- Hot or Not:
- Stop testing now till students return
- For now, will not use people who indicated 4, 5, or 6 on the attractiveness scale because everyone is motivationally relevant to them
- will have a look at the data with Angele and show her how to analyse it

- Value Learning:
- having mega trouble getting Vision Egg to talk to my Mac
- have just tested the program using Asma's laptop, though, and the problem appears to be the code. Will speak to Christian
- New plan: keep working on Python, but also make a version in E-Prime just for the preliminary stuff
- 10 images per category (3 categories), each image seen 5 times each in learning
- testing will use both previously-seen and novel images
- ratings of 30 old and 30 novel images

- SD Review:
- going to meet with Ben next week to get this thing whipped into shape
- plan is to have a super detailed outline ready for Thursday so we all know exactly what we are saying

To do:
- brainstorm three ideas for NSERC
- start VL E-Prime program
- meet with Ben re: SD review
- email Jan at Stonehenge to ask about testing
- try to get Python stuff working (try on Windows side of testing Macs?)

no meeting this week, so just an update

I would just like to say, it is amazing how unmotivated I feel on weeks we don't have a Wednesday meeting. Well, I am still motivated (lord knows I am DYING to collect some Hot or Not data!) but I don't have the same excitement about the research as after I've just spent an hour discussing it with someone who thinks it is equally cool. I guess you can take that as a giant compliment, Mark.

Anyway, things continue. Today I am off to Stonehenge to (hopefully) collect a bit more data from the women. I was sort of worried I'd collect all this and never use it, but Mark seems certain we will use it for SOMETHING, so that's good. It really is a giant pain, so I'd hate to do it all for nothing!

Last week I went to the Stonehenge recognition lunch in an attempt to build some bridges. It was kind of weird - they have a LOT of rules about which I know nothing. It's really like being in a foreign country sometimes, where you have no idea what the etiquette is. However, they fed me a hot dog and red juice, and then I went to their little awards ceremony and that was nice. (Oh, and attended a fire drill, just for authenticity!)

In other news, Ben has cancelled our SD review meeting for this week, so am guessing he hasn't done much. I've collected some information but haven't done much writing. Maybe today or tomorrow - really want to get some kind of draft done, even a rubbish one.

On the plus side, we are in good shape with Hot or Not. Had a big setback last Friday when the images weren't lining up properly and it was a matter of messing with the code rather than the easy non-code E-Prime. I went in on the holiday Monday (dedicated, I know!) and managed to get it working again. The images were also really small, but I have come to the conclusion that they looked full-size on the SurveyMonkey site just due to differences in how they are displayed for the web vs. on E-Prime. I doubled the size of them for the E-Prime and they look fine. I've also gone through and fixed a bunch of images that got cropped incorrectly (looks like they were run through a batch processor at some point), so that is all done.

So currently, Angele is making the practice programs (I am hoping it should be easy enough for her - the images are all selected) and then I need to run through the whole thing and have a look at the data, and then I think we can run. We are having technical issues where I can't get my thumb drive to connect to the testing computer so she can work on it at the lab - very annoying. However, we will get it done.

And finally, should I get a burst of energy this week, I would like to a). do my Python homework as I've fallen way behind from missing a few weeks, and b). sort out the value learning experiment. It may be that we could test both together within an hour, which would be excellent.