GESIS Training Courses
user_jsdisabled
Search

Scientific Coordination

Alisa Remizova

Administrative Coordination

Claudia O'Donovan-Bellante

Please wait...

Data Quality Assessment for Survey Responses: Be Careful of the Careless

About
Location:
Online via Zoom
General Topics:
Data Analysis, Suvey Methodology, Research Data Management
Course Level:
Beginner
Format:
Workshop
Software used:
R
Duration:
2 days
Language:
English
Fees:
Students: 220 €
Academics: 330 €
Commercial: 660 €
 
Keywords
Additional links
Lecturer(s): Matthias Roth, Thomas Knopf

About the lecturer - Matthias Roth

About the lecturer - Thomas Knopf

Course description

In survey research, some of the collected responses might be of low quality. This can have various reasons, such as survey design issues, respondent fatigue, lack of motivation or understanding, carelessness, and response biases. Low-quality responses can threaten the validity and reliability of the research findings and data analysis. Low-quality responses can also affect the effectiveness of statistical analyses, leading to inaccurate interpretations or misleading conclusions. Thus, identifying and addressing low-quality responses is crucial to ensure the integrity and robustness of research outcomes.
 
This is an introductory workshop on detecting and handling quality issues in the data collected from survey research. The primary focus is the quality of responses to the Likert scale questions. After the workshop, participants will have an overview of response quality issues in survey research and know indicators to diagnose and deal with quality issues.
 
Specifically, this workshop covers the following topics:
  •     Introduction to response quality in survey research
  •     Introduction to a set of response quality indicators, both theoretically and practically. In more detail:
    •     Paradata - usage of, for example, timing variables/response latencies
    •    Check items - usage of, for example, directed and undirected attention checks (Bogus items) and similar items
    •   Item responses - usage of three classes of response quality indicators that address outliers, consistency, and response patterns
  •     A discussion and recommendations on how to flag and filter respondents that do not meet quality standards
 
We will provide software demonstrations and hands-on exercises on data quality assessment using R. We will also provide example datasets and R-scripts. The course material will be presented as PDF slides in English. Participants are encouraged to bring their own data to receive the lecturers' recommendations about data quality assessment.
 
Organizational structure of the course
The workshop will include lectures and practical assignments, carried out individually or in small groups and consisting of participants' projects, group discussions, and hands-on exercises with provided cases.Lecturers will be available for individual consultations on participants' projects (on 2nd day), will support work on assignments, and facilitate discussions within group assignments.


Target group

Participants will find the course useful if:
  • they are planning or have already conducted surveys, such as social science surveys, customer surveys, or employee surveys
  • they analyze quantitative data from surveys, which include single items or/and multiple-item scales, and want to improve the data quality
  • they aim to acquire robust skills that are necessary to understand, evaluate, and handle data quality issues in surveys


Learning objectives

By the end of the course, participants will:
  • know data quality issues and concepts
  • know how to evaluate the quality of survey responses and detect and handle low-quality data in the sample
  • gain hands-on experience in data quality assessment in R, using commonly applied data quality indicators in survey research


    • Prerequisites

  • Basic knowledge of survey research terminology (survey mode, questionnaire, measurement instrument, item, type of question format, response options)
  • Basic understanding of descriptive statistics (counts, proportions, means, (co-)variances, correlations)
  • Basic skills in R (RStudio) for data handling and analysis (above all: package handling, data import and data wrangling/inspection of data frames)
    •  
    Software and hardware requirements
  • R version ≥ 4.2, for convenience, ideally in combination with the most recent RStudio version
  • R-packages for data import: (readr, sjmisc etc.); for data wrangling: tidyverse; for data analysis: e.g., psych
  • When bringing your own data, please prepare a wide-format data file (one row per participant)
    • Note: Pay attention that you have sufficient (administrator) rights to install packages on your computer during the workshop!


    Schedule

    Recommended readings