Analyzing data from IEA´s studies with R
Time: 8am-12pm, March 10th Room: Newberry/ Ascot
Objectives of the workshop
a. To provide an introduction to R´s language and environment. This will mainly be oriented to large scale assessment (LSA) datasets. Emphasis will be given to the understanding of how to handle and manipulate data in R.
b.To give participants the opportunity to get familiar with a group of R functions created specifically to analyze studies with complex sampling and assessment designs, i.e., IEA’s Trends in International Mathematics and Science Study (TIMSS), and Progress in International Reading Literacy Study (PIRLS).
c. To give participants the opportunity to expand the functionalities of the IDB Analyzer (a software developed by IEA for the analysis of its datasets) in R. Participants will be encouraged to exploit the flexibility of R to fulfill specific research needs. Moreover, statistical methods not covered by the IDB Analyzer will be introduced (i.e., logistic regression),
d. To give participants an introduction on how to create self-designed graphs with LSA data
Rationale of the workshop
R is an open source statistical programming language which is becoming a standard for statistical analysis in many fields. Its integrated facilities for data manipulation, calculation and graphical display have established R as one of the leading statistical programming software products on academics and research institutions.
Though there are several available software packages that allow the manipulation and analysis of LSA data, R has a variety of strengths which makes it stand out from other commercial options:
- It is a free and open source software
- It is effectively platform independent (i.e., Windows, Mac, Linux, etc.)
- Extensive help resources (i.e., r-project.org)
- Data manipulation oriented environment
- All available functions can be extended or customized depending of users’ needs
- High quality graphics
The workshop will give researchers and students the basic tools to analyze data with complex sampling and assessment designs in R. Firstly, this workshop will provide participants with the necessary tools to reproduce some of IEA IDB Analyzer analysis functionalities. Importantly, the idea of this course is not to provide participants with an alternative to the IDB Analyzer, but to make them familiar on how to take into account the special features of LSA studies (i.e., TIMSS & PIRLS) while analyzing the data in R. Secondly, this workshop seeks the extension of analytical tools available by the IDB Analyzer (i.e., non-linear models). Thirdly, an introduction of how to create self-designed high quality graphics will give participants an extra tool for their analyses.
Target Audience
a. he course is aimed at graduate students and researchers who are interested in the analysis of international large scale assessments in a free, flexible, open source platform.
b. Participants should have:
i. Basic understanding of the characteristics of educational LSA datasets (sampling and assessment designs)
ii. Working knowledge of inferential statistics
iii. Basic knowledge of R
iv. Knowledge of IEA’s IDB Analyzer or similar software would be an advantage
v. Participants should have R version 2.15.1 or latter installed on their computers
Instructional goals
After the training participants will:
a.Have a basic understanding of how to handle and manipulate IEA’s data in R.
i. Importing the data
ii. Exporting the data
b.Become familiar of how to use and customize a number of functions specially created for educational LSA studies for descriptive statistics
i.Frequencies, means, percentage, mode and median
ii.Cross tabulations
c.Learn how to use and customize functions specially created for educational LSA studies for inferential statistics
i.Linear regressions
ii.Logistic regressions
d.Be introduced to a series of important commands to exploit the high quality graph production features of R.
In summary, the course will give participants an overview of how to exploit R´s flexibility while analyzing data with complex designs. The workshop is not intended to teach participants a specific way of doing things, but to introduce them to R´s statistical language and how to deal with data design features while doing any kind of analysis in R.
Planned workshop: schedule and activities
a)An overview of statistical complexities in IEA’s LSA datasets (30 minutes)
b)Preparing and exploring the data in R (45 minutes)
c)Replicating IEA’s IDB Analyzer selected functions in R (30 minutes)
d)Additional analyses not available in the IEA’s IDB-Analyzer (Logistic regression models) (40 minutes)
e)Creating graphs (45 minutes)
B) Instructional staff and institutional affiliation
Course director: Diego Cortes, M.Sc.
Position: Statistician
Institution: IEA Data Processing and Research Center
e-mail: diego.cortes@iea-dpc.de
Plamen Mirazchiyski, PhD
Position: Researcher
Institution: IEA Data Processing and Research Center
e-mail: plamen.mirazchiyski@iea-dpc.de |
|
