Bienvenidos a mi pagina Welcome to my page

Kansas State University KSU Physics Department  REU Program University of Texas at Brownsville UT Physics Department

This page summarizes my research experience during summer 2014 with Dr. Sayre on Physics education. I worked with another REU student, Edward Schenk.

Below you will find my Project Overview, my Research Description, my Research Progress, and my Poster. After all this you will find information About Me. Under the title you would find some useful links.

We conduct a meta-analysis of data from all papers published in PhysRevST-PER, the PERC proceedings, and AJP that measure student performance on the FCI and FMCE in college-level courses (76 papers, ~51k students). The papers' descriptions of teaching methods are coded according to their level of interactive engagement (a lot, some, and none). We augment published data about teaching methods with institutional data such as Carnegie Classification and SAT scores. We statistically determine the effectiveness of different teaching methods as measured by FCI and FMCE gains and mediated by institutional and course factors. In contrast to the landmark 1998 Hake study, a broader distribution of normalized gains occurs in each of traditional and IE classes. Over time, average gains in classes described as "traditional" have increased, suggesting that the definition of traditional teaching has drifted.

During this summer, I learned how to run statistical analysis using R. For the most part we used to do graphs. We also used it to run ANOVAS which is useful to find out if samples within a population are similar or different, and after the anova we run Tukey HSD test to find out which pair of samples are different from each other and which one are the same. In order to run those analyses you have to use a code. The following image shows R when you first open it.

The following lines are a code sample and in front of it you can find a small description of what each line does. Red is the code, black are comments and blue is the result.

Mechanics = read.csv("20140718-ECS-Mechanics.csv")   #This reads the file and sends it to mechanics.

attach(Mechanics)  #It makes objects within the dataframe.

summary(aov(Hake_gain~Test*Teaching_Method))  #Runs an anova with to factors

Df  Sum Sq Mean Sq F value   Pr(>F)

Test                                            1     1.504     1.504    103.80   < 2e-16 ***

Teaching_Method_Shorter          1      4.626     4.626    319.21   < 2e-16 ***

Test:Teaching_Method_Shorter   1       0.331    0.331    22.82    2.39e-06 ***

Residuals                                  465      6.739   0.014

---

Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

19 observations deleted due to missingness

TukeyHSD(aov(Hake_gain~Test*Teaching_Method_Shorter)) # Tukey Test, less than 0.05 means significant different, and greater than 0.05 means they are significant the same.

FMCE:IE-FCI:IE              0.11147326  0.07866423  0.14428229 0.000000

Here are some other graphs that we made that are not on the poster.