| second mtg | third mtg |
fourth mtg | fifth mtg
|
sixth mtg | seventh mtg |
eighth mtg | ninth mtg |
tenth mtg | eleventh |
twelfth | thirteenth
|
fourteenth |fifteenth
|
Ryan Lab Group Meetings
Fall 2013 (2138)
Prior to start of semester
Fall 2013 (2138)
First Week of Fifteen, Mon. 08-26-13
|
back to top |
next |
Fall 2013 (2138)
Second Week of Fifteen, Mon. 09-02-13
Fri., 09/06/13
Met with Sean.
DRM - effect of percieved authority on number of critical lures falsely
recognized after having been missed on a recall test.
Half of
each of the above conditions got a strong suggestion half got no
suggestion.
The result was that both the professor and the student were rated as
highly professional (in other words the manipulation check showed that
the manipulation did not really occur)
False recall rates were almost at ceiling for both the high and low
percieved authority - no significant difference.
Sean will not proceed with that experiment
CP 4.5
Effect on performance on two Learning (no transfer) problems and
two transfer problems
The Learning problems were same surface features and same unknown as
training
The transfer problems were both new surface features one same unknown,
one new unknown
of Backwords Fading vs No Backwords Fading (a replication of CP 4)
and Disc and Mix for training and Travel for transfer
vs.
Disc and Travel for training and Mixture for transfer
Another dependent measure is - they were asked to provide the
underlying equation
25 subjects so far. In a 2 x 2
But the 25 subjects may not be balanced in the cells. Sean will check
the coding.
Then will start running again.
|
previous |
back to top | next |
Fall 2013 (2138)
Third Week of Fifteen, Mon. 09-09-13
Thurs.,
09/12/13
Met with Dan Reynolds who had worked on the Spring 2012
Motivation study in Stats. He pointed me to the raw data. However, we
need to contact Dave Kile to get more info on what the raw data means.
Dan will contact Dave.
I explained the Center for Open Science - the Reproducibility Project -
the Framework for Open Science.
Once I hear back from Dave and Dan, I'll decide what to have them do,
and what I can use the new volunteers to do.
|
previous |
back to top | next
|
Fall 2013 (2138)
Fourth Week of Fifteen, Mon. 09-16-13
Thurs., 09/1913
Met with Sean about CP 4.5 - He found some miscoded data (wrong
condition) which he fixed. After fixing the coding, he got the
following means of a 2 (Backwards fading or not) by 2 (Trained w/
Disc&Mix vs. Trained w/ Disc&Travel). But first, note that the
N's are small and unbalanced. After the N's are the 2 x 2's for Total
on posttest, then individual posttest problems. None of the analyses
were done as anovas, and, therefore, no covariates were used (not worth
all that effort with these small and unbalanced N's). However, there
are two tailed t tests comparing the marginal means for the Training
problems factor.
N's
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
9
6
Discount &
Travel
3
7
Total on the posttest as %
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
83.33
62.50 .7321
p = .082 two tailed t
Discount &
Travel
25.00
53.57 .5000
Find final / non-transfer as %
(procedure same as train/same surf as train - discount)
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
89.00
83.00
Discount &
Travel
67.00
86.00
Find final / non-transfer as %
(procedure same as train/same other surf as train - mixture or travel)
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
89.00
50.00
Discount &
Travel
33.00
71.00
Find final / transfer (to new surf) as %
(procedure same as train/surf opposite of whether they got Mix or Trav
at training)
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
100.00
83.00 .93
p = .007 two tailed t
Discount &
Travel
0.00
57.00 .45
MOST DISTANT TRANSFER
Find initial / transfer (to new surf) as %
(PROC DIFF FROM TRAINING/surf opposite of whether they got Mix or Trav
at training)
Training problems
Backwards
Fading
No Backwards Fading
Discount &
Mixture
56.00
33.00 .43
p = .066 two tailed t
Discount &
Travel
0.00
0.00 .09
Sean would like to change the experiment somewhat and run again. The
change would be that on the posttest, but not the pre-test, there would
be four additional test problems. So that instead of receiving only one
of each type of test problem, they get two. Then he will try to run
more subjects.
|
previous |
back to top | next
|
Fall 2013 (2138)
Fifth Week of Fifteen, Mon. 09-23-13
|
previous |
back to top | next |
Fall 2013 (2138)
Sixth Week of Fifteen, Mon. 09-30-13
|
previous |
back to top | next |
Fall 2013 (2138)
Seventh Week of Fifteen, Mon. 10-07-13
|
previous |
back to top | next
|
Fall 2013 (2138)
Eighth Week of Fifteen, Mon. 10-14-13
Thurs. 10/17/13
Derek and Erik are still working on the KLI paper. We discussed the
meaning of "Knowledge Components".
Talked about the CET meeting. There will be future meetings to try to
get faculty for various departments working together on "The Science of
Learning"
Also, PASSHE apparently is putting together an NSF grant on the Science
of Learning for all 14 schools. Dr. Paul Ache, the Chair of the Math
Dept. here is working on that. I'll contact him to find out more.
|
previous |
back to top | next
|
Fall 2013 (2138)
Ninth Week of Fifteen, Mon. 10-21-13
|
previous |
back to top | next |
Fall 2013 (2138)
Tenth Week of Fifteen, Mon. 10-28-13
Thurs. 10/31/13
Meeting with Erik. Last time I gave Derek and Erik the Interleaving
paper in progress. I briefly described the 6 studies, and asked them to
try to use the KLI paper to come up with a better way to apply the KLI
framework to stats. Since Erik has not yet had Stats, it would be hard
for him to apply the KLI framework to analyzing that knowledge. But, he
could be taking Stats next semester. So, next semester, he (and whoever
else I can get to work on it) could be working up a study on how to
apply the KLI framework to Stats that we could run in Fall 2014.
I sent Chi & Bassok's 1989 book chapter on Learning From Examples
Via Self-Explanation to Derek and Erik as the next paper to summarize.
|
previous |
back to top | next |
Fall 2013 (2138)
Eleventh Week of Fifteen, Mon. 11-04-13
|
previous |
back to top | next |
Fall 2013 (2138)
Twelfth Week of Fifteen, Mon. 11-11-13
Thurs. 11/14/13
Met with Erik Schwambach and Derek Rickards.
Chi and Bassok 1989. Erik reports that they used physics problems
involving blocks and inclined planes because people tend to think about
what they can see, and not about gravity or force, which they can't
see. The book chapter says that the key type of Self Explanation that
helps students are those that relate some solution step in the example
to some abstract principle that they learned from the text in such a
way that the student now notices that the abstract principle provides
the rationale for the step.
One thing we need to do is to get more info out of that book chapter to
see if we can figure out what kinds of worked examples we could present
in a statistics class in such a way that they will elicit those kinds
of SE's described above.
Also, as shown below, Derek analyzed the "which stat procedure for which research
situation" task into KC's ala the KLI framework.
=============
Feed Back and Interleaved Examples vs.
KLI Relations
*What Knowledge Components (KC) are
involved in Blocking and Interleaving Instructional Events (IE)?
*What Assessment Events can be used to
measure the effect of learning these Instructional Events?
Learning Events that would develop
in specifying what statistical test to use.
Understanding
and Sense-making: Categorize when different tests are applicable
to each situation. Ex. # of Independent variables (1 or 2) and # of
Dependent variables (1 or 2).
Induction
and Refinement: What are the types of Independent variable and
Dependent variable conditions? Ex. What kind of scale is it?
(Nominal, ordinal, interval).
Memory and
Fluency Building: Most expansive LE… What steps are taken to
solve each test? What situation to use each test? (Formulated from
numbers 1 and 2)
Guidelines for “what
statistical analysis to use” was developed by James D. Leeper,
Ph.D.
http://www.ats.ucla.edu/stat/mult_pkg/whatstat/
Next step is to…
-- Establish what Knowledge
Components are involved in Blocking and Interleaving styles.
--
Determine what Instructional Events best suite these Learning Events
and Knowledge Components.
=============
We will examine that
analysis further with an eye to using it to develop some instructional
techniques to test. Next semester, Erik will be taking Stats, so he
will have a better idea
of how to provide input into developing the instructional methods to
test (Derek already had Stats).
I reminded Erik and Derek that this semester we just have next week
(then the following week is Thanksgiving) and the last week of the
semester to meet, but that we need to continue this next semester. So
it's important that they continue with this.
|
previous |
back to top | next |
Fall 2013 (2138)
Thirteenth Week of Fifteen, Mon. 11-18-13
Thurs., 11/21/13
Met with Erik Schwambach, Derek Rickards, and (new) Troy
Vannucchi. We discussed how to implement the ideas from Chi and Bassok
1989 in a stats learning experiment. Derek suggested the idea of giving
subjects instructions first in the "principles" : "If there are only
two groups of data, then use a t test, whereas if there are more than
two groups use an anova" and "If the groups of data come from different
subjects, then use an independent measures test, whereas if they come
from the same subjects measured more than once, then use a repeated
measures test". Then, give them examples, require them to state
which principle they should apply and why, and follow that with
feedback. Derek, Erik, and Troy will all try writing up the skeleton of
the procedure for such an experiment with an eye to deciding how to do
the control group.
I asked them to consider how the experiment we produce would be
connected to the previous literature. We want to make sure we can
justify that the experiment was based on the previous literature, not
just an idea out of thin air. We especially want to be able to connect
it to the KLI paper, as well as perhaps, the Chi and Bassok.
Also, I gave them the title of the Chi, DeLeeuw, Chiu, and Lavancher (1994) paper so that they could get it from PsychINFO.
Finally, I reminded them that we will meet again two weeks from now
(after Thanksgiving), and then, whereever we are at that point, we want
to be able to pick up right from there and continue when the Spring
semester starts.
|
previous |
back to top | next |
Fall 2013 (2138)
Fourteenth Week of Fifteen, Mon. 11-25-13
|
previous |
back to top |
Fall 2013 (2138)
Fifteenth Week of Fifteen, Mon. 12-02-13
Thurs., 12/05/13
Met with Erik Schwambach and Troy Vannucci.
I showed them and emailed them copies of some materials from the
Feedback and Interleaving experiment (Spring 2011 and Fall 2011). I
sent the script for the training and immediate test, and the subjects'
booklets for the No Feedback and the Feedback conditions (both blocked
- no interleaving - both order 1).
They will use those to try to write up some materials for an experiment along the lines suggested by Derek on 11/21/13.
In the meantime, I will continue to see if, and how, the KLI framework can be used to motivate such an experiment.
Erik will be around during the winter break and will be working part
time. Troy lives an hour and a half away. Usually he works, but not
sure yet for this winter. We should keep in touch over the break and be
ready to start right at the beginning of the Spring semester.
Fall 2013 (2138)
After the semester ended