University of Connecticut University of UC Title Fallback Connecticut

Workshops

SCS is pleased to announce two introductory workshops that cover survey design and power analysis / sample size calculation!

Where: Austin (AUST) 105

When: Wednesday, October 26, 2016; Wednesday, November 16, 2016
Session 1: Power Analysis and Sample Size Calculation (10/26 3:30PM – 5:30PM)
Session 2: Survey Design (11/16 3:30PM –5:30PM)

Who: Any UConn or UCHC faculty, students, or affiliated persons are welcome to attend.

Registration: http://merlot.stat.uconn.edu/www/consulting/workshops/register.php

Refreshments will be served in AUST 326 at 3:00pm before each session. Participants can sign up for one or both of the sessions. Registration will be closed whenever the cap number (100 for each session) is reached or by Oct 19, 2016 for session 1 and Nov 9, 2016 for session 2.

Session 1 (10/26): Introduction to Statistical Power Analysis (Sample Size Calculation)

Session 1 (10/26): Introduction to Statistical Power Analysis (Sample Size Calculation)
This workshop shall focus on the basics of statistical power analysis, in particular the calculation of required sample size for an experiment. Power analysis is an extremely important component in the design of efficient experiments, ensuring the detection of important effects with a high probability. A broad overview of the components of power analysis shall be the starting point of this workshop, after which specific statistical problems shall be tackled using case studies and live demonstrations using the freely available G*Power software. For each case study, participants will be taken step-by-step from the initial problem statement to the final power calculation results, considering important design and computational guidelines along the way. At the end of the workshop, participants will have a broad understanding of power analysis and will have learnt how to solve the most common power calculation problems using G*Power. This workshop consists of 4 sub-sessions: 1) Overview of Power Analysis; 2) Power for difference of means; 3) Power for proportions. There will be a thorough overview of each topic and a hands on, live demonstration of various case studies

Prerequisites: An introductory course in undergraduate statistics and familiarity with the basic concepts of statistical hypothesis testing are desirable. Participants are encouraged to bring their own laptop computers to the workshop and install the freely available G*Power software (http://gpower.hhu.de/).

Session 2 (11/16): Survey Design

This workshop will give a general overview of survey design. We will cover the major concepts needed for constructing a survey from scratch starting from defining research questions up to analyzing the data collected. We will discuss a step-by-step procedure for designing surveys and common errors or issues which may arise from deeper statistical concepts to question phrasing. We will also discuss some useful steps to take before performing statistical analysis and some common analysis techniques, along with demonstrations. The work shop will finish with a discussion of interrater reliability which can help assess the validity of your data and refine surveys by determining if the appropriate scales have been chosen for particular measure. All topics will be illustrated with examples.

Prerequisites: Experience with ANOVA will be useful but no prior statistical knowledge is required for the bulk of the workshop.