=================================================================== 701 Project lecture slots (MWF 130-220 PHA19D; 13 students; 7 projects; 9 lectures) Wed Nov 1 Lori Holt - language acquisition through eye tracking Fri Nov 10 Cliff Davidson - inconv truth project Wed Nov 15 Alison Barth - biochemical markers of neuron interconnections with learning Thu Nov 16 Bill Eddy on Ferret visual system neuron growth [special time/place] Fri Nov 17 Mon Nov 20 Mon Nov 27 Judy Cole - analysis and design of alum survey(s) Wed Nov 29 Jack Mostow & Ayorkor Mills-Tettey - implementing tutoring system in Ghana Fri Dec 1 Marc Sommer - connections between the brainstem and frontal cortex Mon Dec 4 Ken Koedinger - measures of learning in online tutoring systems lori holt jack mostow assistments judy cole - alum surveys - still fishing PREFERENCES: * Zhanwu Liu (zhanwul@stat.cmu.edu) and Yi Jiang (yijiang@stat.cmu.edu): Judy Cole * Nathaniel: First: Ghana tutoring project Second: Language acquisition through eye tracking Third: inconv truth project * April and Nathaniel: Ghana * Han Liu 1. Judy Cole 2. Cliff Davidson * Elise Olson 1. Ghana 2. Koedinger 3. Cliff Davidson * Joey * Gaia Bellone & Shuhei Okumura First choice : Marc Sommer - connections between the brainstem and frontal cortex Second choice : Lori Holt - language acquisition through eye tracking Third choice : Ken Koedinger - measures of learning in online tutoring systems * Gabrielle Fijas and (gary?) Bill Lori * Wasik (work with Crossett) 1. sommer 2. barth 3. koedinger * Crossett (work with Wasik) 1. sommer 2. barth 3. holt =================================================================== 401 Project lecture slots (TR 300-420; SH 219; ca 24 students, ca 7 projects; 7-8 possible meetings) Intended to do them within 20 min at the beginning of class. Thu Oct 26 Inferences; PROJECTS Tue Oct 31 (Extra SS; multicolin) Thu Nov 2 (Poly Reg & Interactions) Tue Nov 7 (reg with cat predictors) Thu Nov 9 (cat & cont predictors Tue Nov 14 (model selection) Thu Nov 16 (model checking; influence) Tue Nov 21 (projects; lab meeting) [peter freeman - two projects] Thu Oct 26 Rebecca Nugent - PFT and Obesity/Kids Tue Oct 31 Thu Nov 2 Lori Holt - *last* 20 min of class - needs projector for laptop Tue Nov 7 Jacob/Daniel (401 students) - Pedestrian Safety Data Thu Nov 9 Cliff (parking survey) Tue Nov 14 Joel Tarr - Historical property surveys Thu Nov 16 Ken Koedinger Tue Nov 21 assistments =================================================================== Totals: 402 726 Jim / Pitt usability x dead Reb / Pulm Fcn Test x date for 402 Reb / Obesity - Kids x date for 402 Reb-Tanz / P Freeman x date for 402 Jim / Joel-Howard [waiting] Tanzy / Steve dead Tanzy / Tarr / WPA ? ? [asking tanzy for detail] BJ / Alison Barth x date for 726 BJ / Lori Holt x x dates for both BJ / Assistments x ? dates for both BJ / Jack Mostow x date for 726 BJ / Alum Survey x [waiting] BJ / Cliff Davidson x x dates for both --- BJ / Mitch Small in EPP ASKED BJ / John Miller in SDS BJ / PIER BJ / Ambrose / Eberly ASKED BJ / Glorianna StClair BJ / Bobby Klatzky BJ / LRDC / IQA -- dead BJ / Bill 2-3 projects =================================================================== PROJECT POSSIBILITIES Jim - checking with his colleague in finance who will be speaking at uncoming research tea Jim - Jay / Group at Pitt who do usability studies of sr mobility devices Rebecca - Pulmonary Function Test Rebecca - Obesity in Kids & TV Rebecca - Peter Freeman / Astronomy Data Rebecca - Joel & Howard summer projects? Tanzy - Joel Tarr / pedestrians project Tanzy - Steve / CDC data ------------------------------------------------------------------- BJ* - Alison Barth (Kass email contact has occurred) send inquiry Wed Oct 11 21:53:04 2006 She writes, All sounds good. I have a project involving the properties of small spontaneous excitatory and inhibitory post-synaptic currents in neocortical neurons from three different groups. We have collected data representing thousands of these events, and I think that they can be classified into different groups according to rise and decay times of the events (i.e. fast rise and slow decay, slow rise and slow decay, fast rise and fast decay, all factors versus amplitude of the event). The large sample size will make this analysis easier, and I already have some idea of what we might find. Let me know what date you'd like me to present. Sooner is better. READY TO PRESENT -------------------------------------------------------------------- BJ* - Kass suggests Lori Holt inquiry Wed Oct 11 22:10:10 2006 meet weds 11 am... READY TO PRESENT -------------------------------------------------------------------- BJ*** - Assistments Explore relationships between - assistment scores - MCAS (state exam) score - MAPS (another benchmarking system I think) (MAPS is a computer adaptive testing program from www.nwea.org) Does learning obey power or expon law? (basically logistic regr with error as response...) - what alternative to logistic reg if power law applies? - maybe use the old geom data instead of assistments? Same question, with assistance score (approx, errs+hints), rather than errors alone. - is our current assistance score "optimally constructed?" PROBABLY READY TO PRESENT -- NEED A PERSON 402 ------------------------------------------------------------------- BJ* - Jack Mostow / Nathaniel "is this real" email Wed Oct 11 21:48:14 2006 READY TO PRESENT ------------------------------------------------------------------- BJ* - John Lehoczky / Judy Cole / analyze one alumni survey and design another sent query Wed Oct 11 21:31:25 2006 call MTW am direct line 412-268-6286 [full number!] Tue Oct 17 11:41:18 2006 went to john for a reality check feeling that the marketing survey wasn't that helpful for whole-pop understanding (e.g. only sent ot email addresses) how about let's do a statistically valid rand sample maybe even creating a panel of alumni for repeated sampling that would be statistically valid how do you know what the right strata are to create that random sample sick an MS student on it: how do you make better choices when the data you have are inadequate For stat pract: need a little data analysis and then can suggest survey design issues For ADA: Do the full design implement and analyze process starting THIS SUMMER SOME FOLLOWUP NEEDED ------------------------------------------------------------------- BJ* - Jay / Cliff Davidson / detecting air pollution from Bellefield plant in Schenley Hollow up on flagstaff hill; now have NOAA cloud cieling data; nanjun archives data inquiry Wed Oct 11 22:36:56 2006 arr mtg for Tue Oct 24, 4pm ------------------------------------------------------------------ BJ - Jay / Mitch Small in EPP BJ - Jay / John Miller in SDS BJ - PIER or other psych folks BJ - Susan Ambrose / Eberly Ctr BJ - Glorianna StClair / Libraries BJ - Kass suggests Bobby Klatzky BJ - LRDC / IQA BJ - Bill can support 2-3 projects. Do email coord.