miércoles, 3 de agosto de 2016

Presentation: Math, Statistics, R programming, online distance education and other topics covered in this blog.

This blog- or rather its predecessor Google site (now defunct)- originally began in 2009 as an auxiliary tool to communicate on-line with my students in the Universidad Nacional Abierta of Venezuela (UNA), a distance education university where I teach math and statistics. The original purpose of that site was to post grades, some bibliography and model exams as well as some announcements- all in all a very modest and limited educational application of information technology. In time, I began to publish some entries dealing with frequently asked questions, but mostly as a time-saving device for me, as I would point my students to this content.

I soon saw that the Google site platform was not so well suited for this purpose and so moved this site to the present blogger platform. I would still continue to use the new blogger site fundamentally as a repository and to post various clarifications of math and statistics related topics that came up in the subjects I teach. Throughout this time, I had become passionate about R programming and its use to learn math and statistics (I myself having benefited greatly in this regard), and so I would occasionally throw in a post or two with examples of R programming to broaden my student's horizons, although still keeping in line with the course contents. Alas, despite these entries (some with very original content in my opinion), my readership did not grow as I intended and the blog very much lay dormant until some time recently.

Although somewhat disillusioned by my blog's traffic and the uninterested reception of these posts among my students, my interest in R programming and its applications in college-level mathematics and statistics education did not waver. For example, I developed an R library called estUNA oriented towards students of statistics courses that could be used on RWeb servers on-line. This would permit students to accomplish the computational activities of the statistics projects and use the data sets issued for the UNA courses each semester, on-line and without having to download and install R on their computers.

The intent of estUNA was to make R use and its characeristic console work flow accessible to students with no programming background and very little computer skills by easy to understand function names in Spanish, by including the datasets directly in the library, thereby relieving the student of complicated data conversion and loading procedures, and even by allowing the student to use R and estUNA on a browser on-line without having to download and install R, let alone additional R packages. Throughout these years, I have blogged about this "R solution" for students in my blog, sometimes with detailed examples of its use in various statistical analyses that are topics dealt with in UNA statistics courses.

Another interesting tool I developed in R is HEVASU. Before going into that, I must very briefly explain the Universidad Nacional Abierta's educational and organizational model. The UNA is a nationwide university distance education institution in which course contents and plans are centralized and created by specialized personnel in Caracas. The actual interactions between students and the institution takes place in local and regional centers across the country where students enroll and consult with professors like myself, who act more as facilitators than as teachers in regular classrooms.

In this distance educational model, students are given books (in paper) and detailed course plans including schedules for presential exams. Thereafter, most of the instruction is self instruction and we teacher/facilitators intervene when the student calls upon us for consultation.  If this seems like a blast to the past for some readers acquainted with MOOCs and more modern ideas of distance education, eLearning, and the like- well, in my opinion it is too, but that is another topic I'll be blogging about in future posts.     

HEVASU was originally an R application I made for myself to simplify the creation of student lists every semester in Excel format for each of my subjects and other associated bookkeeping tasks which makes part of the drudgery of every teacher's profession. Every trimester, we have to fill out activity reports with numbers of how many students came to consultation classified by subjects and undergraduate programs, numbers of graded exams and projects, etc. The original intent of HEVASU was to automate the creation of these reports in Word and pdf formats. In time, I added other functionalities like automated creation of certificates for the UNA introductory course by integrating with LaTeX, all of this in a GUI application done solely in R.

Surely this may strike you as a strange application to do in R, being a language oriented mainly towards statistical and quantitative analysis. Why R and not some other language like Python? Well, it had begun to dawn on me that all this data, collected semester after semester with detailed information about when and which students come to me for consultation on what subjects, when and who comes to take presential exams, what course objectives do they master, etc. was a valuable data to mine and analyze. Surely, I could have still done HEVASU in some other language and the actual statistical analyses in R, but my idea then was to integrate all that into a single application. Nevertheless, in the process I learnt a great deal about GUI programming in R, pdf to text conversion using regular expressions and automated report and certificates generation all which are aspects I'll be blogging about too.

I had by then taken a few MOOCs myself on the topic of Data Analysis and Machine Learning (including an excellent one by Professor Abu Mustafa of Caltech). Additionally, the massive aspect of MOOCs themselves and their potentiality for using student micro-events data to derive powerful educational insights, not to mention the revolutionary global democratization of knowledge that these courses made possible fueled my interest in alternative and emergent paradigms of distance education and my hunger for large data sets on which I could practice these newly acquired techniques. The intent of HEVASU was to produce these large data sets.

This led me to another idea: why not implement formative, quiz-type evaluations on-line to tap into another source of big data? What are the contents and question items most difficult to students about which I should post entries on my blog? What are their patterns for engaging in these on-line quizzes and how does that relate to their eventual success (or failure) in the course? To begin, I chose Mathematics 1, a common course to various undergraduate programs that I believed was critical in student prosecution. I sculpted the quiz generator myself using Javascript and integrated it into this blog.

The on-line quiz system is currently available  here  and if you want to try it out, use my dummy user ID - 12345678 and select the learning objective for the quiz (currently only the first five). After hitting the submit button you'll see a quiz with five multiple-choice questions. After answering (or not) these questions the student will receive detailed feedback on their errors and where they must look in the book for similar problems or clarifications.

This on-line quiz tool significantly boosted traffic to my site. I must say that it is unique in my University and as far as I'm aware of, there's nothing like it even on the Moodle platforms that the UNA's central level is creating for some courses. Besides being of value to my students' learning process, for me it represents a possibility for creating yet another large dataset. All data on which individuals take the quizzes, when, what questions came up, for which objective, what were the answers the student gave for those questions, gets collected real-time in a Google Doc Sheet via Google Forms integration with my site. I'll be blogging here about the process of creating this on-line quiz tool too.

Although my traffic has increased dramatically, I must say I'm still not happy with my blog. Recently I've come to the realization that blogging about my personal experiences and continuing learning process in using computer technology for instructional and educational purposes, specifically in a distance education environment might be of interest to a wider global audience. To do that, I had to broaden my blog to make it bilingual. While still tending to my UNA public in Spanish, I will now also write to an English-speaking audience.

I realize that all this story about how I basically crafted my own technological solutions to instructional problems by a process of trial and error may strike you as a re-invention of the wheel. Maybe in other universities or educational institutions (distance or not), the computer tools are already created and there's not much for the facilitator to do but to apply them.

Even so, this could be of value to you because: 1) It proves that an you as an individual, without having a whole IT department at your disposal, can create this using commonly available and free technology, 2) this allows you to tap into the data that you and your students generate, 3) you can tailor these tools to your particular educational setting and 4) this will enable people, both teachers and students, in underdeveloped countries to become actors in global knowledge economies. But my approach comes at a cost- you have to be willing to learn a little bit more beyond office applications use. You have to be willing to learn how to code. Nonetheless I assure you your efforts will be paid off.

Of course, I'll also be blogging to students, presenting various case studies in math, statistics and even other subject areas with a computational approach (using R) which they themselves can explore and engage in, or simply blogging about math and statistics topics which commonly pose difficulties to college students. I'll also be blogging about data visualization topics and the use of state of the art quantitative techniques available in the extensive R package system as I come into contact and begin to learn about these myself. This is quite an ambitious and formidable blogging program, to be sure. So when does this blog begin? It begins right now!

domingo, 31 de julio de 2016

¿En qué consiste la inferencia estadística? Protocolo de inferencia estadística - parte 1

La primera parcial de la asignatura de Estadística Aplicada (746) contempla los primeros cuatro objetivos, todos relacionados con el tema de inferencia estadística. Bien sea si realizamos inferencia por medio de intervalos de confianza (objetivo 1), contraste de hipótesis de una población (objetivo 2), contraste de hipótesis de dos poblaciones (objetivo 3) o tests de bondad de ajuste o con tablas de contingencia utilizando estadísticos chi-cuadrado (objetivo 4), hay un procedimiento (protocolo) que debemos seguir para resolver los problemas de inferencia estadística que se presentan en la evaluación de estos objetivos. Continuen leyendo más abajo para ver de que se trata.

inferencia.gif

lunes, 25 de julio de 2016

Trabajo práctico de la 745, lapso 2016-1

Como seguramente ustedes saben, los trabajos prácticos de estadística pasaron a ser evaluaciones formativas desde hace algunos semestres. A partir de entonces, el objetivo 1 de Estadística General (745) se evalúa en la primera parcial y en la integral. Aunque ya no es necesario entregar un trabajo práctico cómo tal ni recibirá una calificación departe mía, yo les recomendaría realizar las actividades contempladas en el enunciado del trabajo, que fueron publicadas aqui. La razón de mi recomendación es porque en este semestre, las actividades del trabajo práctico contemplan la elaboración de diagramas de caja y de torta. Mi olfato docente me obliga a avisarles sobre esto e invitarlos a que continúen leyendo esta entrada, donde realizaremos juntos las actividades propuestas, no vaya a ser que las gráficas de tortas y de caja los agarren desprevenidos este semestre...


jueves, 21 de julio de 2016

Métodos de Redondeo (Objetivo 1 de Matemáticas I)

En esta entrada del blog voy a abordar el tema del redondeo, que es contenido del objetivo 1 de Matemáticas I referente al estudio de los números racionales, sus operaciones y propiedades. Pienso que este tema, a pesar de su aparente sencillez, merece una revisión porque, cómo les voy a contar, hasta hace algún tiempo yo mismo tenia una concepción errónea del método de redondeo.


martes, 19 de julio de 2016

Trabajos de Estadística (738/748, 745 y 746) Lapso 2016-1

Se ha publicado los enunciados para los trabajos prácticos de las siguientes materias: 738/748, 745 y 746. Ante todo permítanme aclararles que estos trabajos prácticos son actividades de evaluación formativa y como tal, no tienen fecha de entrega ni ponderación en la calificación final de la materia. Aún así, les recomiendo revisar los enunciados e intentar realizar las actividades que allí se piden. Los enlaces a las carpetas comprimidas con la data y los enunciados en ciberesquina se dan a continuación:

738/748
745
746

Trataré en próximas entradas de abordar las actividades contempladas en esos trabajos. Por los momentos, les recuerdo que pueden utilizar el lenguaje R para realizar las actividades ustedes mismos. En una página fija de este blog se da información para aquellos interesados en usar R. En este mismo blog he publicado video tutoriales sobre cómo instalar R y la librería estUNA así como un ejemplo del empleo de técnicas de estadística descriptiva. También encontrarán muchas entradas ilustrando distintos métodos de estadística en este programa.

Para este semestre, la data se encuentra en el siguiente data frame de estUNA: d20161

Si te gustó o te pareció útil este contenido, compártelo en las redes sociales y dale tu voto positivo en el botón "me gusta" de G+, para que otros puedan encontrar el contenido también.

miércoles, 15 de junio de 2016

Talleres para el semestre 2016-1

Matemática I (175-176-177)


Taller 1Inducción25/06/20168am-12pm
Taller 2Unidades 1 y 202/07/20168am-12am
Taller 3Unidades 3 y 409/07/20162-5pm
Taller 4Unidades 5 y 616/07/20168am-12pm
Taller 5Unidad 702/09/20168am (sede UNA)
Taller 603/09/20168am (sede UNA)


Los talleres de Matemática I serán dictados en el Trujillo. Estén atentos a la publicación de fechas de los siguientes talleres.

Estadística General (745)


Taller 1Objetivos 1-416/07/20169am-12pm
Taller 2Objetivos 5-817/09/20169am-12pm
Taller 3Objetivos 1-822/10/20169am-12pm
Taller 4Objetivos 1-805/11/20169am-12am

Estadística Aplicada (746)


Taller 1Objetivos 1-423/07/20169am-12pm
Taller 2Objetivos 5-824/09/20169am-12pm
Taller 3Objetivos 1-829/10/20169am-12pm
Taller 4Objetivos 1-812/11/20169am-12pm

miércoles, 1 de junio de 2016

Publicación de objetivos logrados

A partir de este semestre 2016-1, la consulta de los objetivos logrados en las materias que asesoro se hará de una manera distinta a como se venia haciendo. Anteriormente, el estudiante navegaba hacia la página de la asignatura cuyos resultados de evaluación quería consultar (ej. Matemática I, Estadística General, etc.) y hacía clic sobre el enlace de objetivos logrados, desde donde se abría una ventana con los objetivos logrados de la nómina completa de estudiantes para esa asignatura.

domingo, 27 de diciembre de 2015

Matemática I - Autoevaluación On-line

Como asesor del área matemática de la Universidad Nacional Abierta, he observado con preocupación que, semestre tras semestre, los índices de prosecución y de aprobación en Matemáticas I (códigos 175, 176, 176) son muy bajos. Creo que los colegas que asesoran esta materia en otras unidades de apoyo o centros locales de la UNA comparten mi inquietud. Matemática I es una asignatura de Estudios Generales que se cursa en casi todas las carreras ofertadas en la UNA, por lo cual no sería exagerado decir que cualquier problemática del estudiante con la asignatura eventualmente genera un cuello de botella para la prosecución estudiantil en nuestra universidad. A continuación esbozo algunas observaciones referentes a cada código de esta asignatura

viernes, 11 de septiembre de 2015

Regresión Lineal, semestre 2015-1

Quisiera en esta entrada aclarar algunas cosas sobre la regresión lineal a la luz de los (deplorables) resultados de la segunda prueba parcial de la 746 de este semestre. Si han presentado esta prueba y no lograron el objetivo 5 o el 6, lean a continuación para ver porqué.

Comencemos repasando lo que es un modelo de regresión lineal:

\[
Y=\beta_0 + \beta_1 X_1 + \beta_2 X_2 + \beta_d X_d + \varepsilon
\]

martes, 30 de junio de 2015

Argumentum ad hominem

El argumentum ad hominem es una falacia lógica que consiste en ataques personales al adversario en un debate para pretender hacer ver que su posición es falsa. Algo de eso ha sucedido en estos últimos días en este blog. Es verdaderamente fastidioso tener que desviar mi energía mental, que pudiese utilizar para cosas mucho más provechosas como ampliar el contenido educativo de este blog, por citar un ejemplo. Pero lo hago, en primer lugar para defenderme de los ataques y en segundo lugar, porque creo que en el fondo, es función del docente formar, no solo impartir contenidos (de matemática en este caso). En tercer lugar, ya que algunos estudiantes han invertido tanta energía mental en averiguar mi vida por internet y darme una importancia que no tengo, espero retribuirles la atención brindada...