The Impact of Mixed Modes on Multiple Types of Measurement Error
Mixed mode designs are becoming standard in the collection of survey data. Despite this, there are still unknowns regarding how mode (e.g., Web) or mode design (e.g., sequential mixed mode) impacts measurement error. Previous research has been limited by the confounding of selection and measurement...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
European Survey Research Association
2020-04-01
|
| Series: | Survey Research Methods |
| Subjects: | |
| Online Access: | https://ojs.ub.uni-konstanz.de/srm/article/view/7450 |
| _version_ | 1811319943981957120 |
|---|---|
| author | Alexandru Cernat Joseph Sakshaug |
| author_facet | Alexandru Cernat Joseph Sakshaug |
| author_sort | Alexandru Cernat |
| collection | DOAJ |
| description |
Mixed mode designs are becoming standard in the collection of survey data. Despite this, there are still unknowns regarding how mode (e.g., Web) or mode design (e.g., sequential mixed mode) impacts measurement error. Previous research has been limited by the confounding of selection and measurement mode effects and the investigation of only one type of measurement error at a time. In this paper, we use three waves of the Understanding Society Innovation Panel to investigate whether single-mode versus sequential mixed-mode and Web versus face-to-face modes have different impacts on measurement error. We make use of a quasi-experimental design that randomly allocated respondents to either a unimode face-to-face interview or a sequential mixed-mode (Web and face-to-face) design. Through this design, we implement a new multitrait-multierror model that estimates social desirability, acquiescence, and method effects simultaneously. The results show no differences in measurement error between single modes and mode designs with respect to acquiescence and method effect but some differences are found for social desirability. We discuss the practical implications of these findings and their possible causes in conclusion.
|
| first_indexed | 2024-04-13T12:52:03Z |
| format | Article |
| id | doaj.art-5edd82d913a847e88d46085c33739855 |
| institution | Directory Open Access Journal |
| issn | 1864-3361 |
| language | English |
| last_indexed | 2024-04-13T12:52:03Z |
| publishDate | 2020-04-01 |
| publisher | European Survey Research Association |
| record_format | Article |
| series | Survey Research Methods |
| spelling | doaj.art-5edd82d913a847e88d46085c337398552022-12-22T02:46:11ZengEuropean Survey Research AssociationSurvey Research Methods1864-33612020-04-0114110.18148/srm/2020.v14i1.7450The Impact of Mixed Modes on Multiple Types of Measurement ErrorAlexandru Cernat0Joseph Sakshaug1University of ManchesterInstitute for Employment Research and University of Mannheim Mixed mode designs are becoming standard in the collection of survey data. Despite this, there are still unknowns regarding how mode (e.g., Web) or mode design (e.g., sequential mixed mode) impacts measurement error. Previous research has been limited by the confounding of selection and measurement mode effects and the investigation of only one type of measurement error at a time. In this paper, we use three waves of the Understanding Society Innovation Panel to investigate whether single-mode versus sequential mixed-mode and Web versus face-to-face modes have different impacts on measurement error. We make use of a quasi-experimental design that randomly allocated respondents to either a unimode face-to-face interview or a sequential mixed-mode (Web and face-to-face) design. Through this design, we implement a new multitrait-multierror model that estimates social desirability, acquiescence, and method effects simultaneously. The results show no differences in measurement error between single modes and mode designs with respect to acquiescence and method effect but some differences are found for social desirability. We discuss the practical implications of these findings and their possible causes in conclusion. https://ojs.ub.uni-konstanz.de/srm/article/view/7450multitrait-multierrorexperimental designmeasurement errormixed-mode design |
| spellingShingle | Alexandru Cernat Joseph Sakshaug The Impact of Mixed Modes on Multiple Types of Measurement Error Survey Research Methods multitrait-multierror experimental design measurement error mixed-mode design |
| title | The Impact of Mixed Modes on Multiple Types of Measurement Error |
| title_full | The Impact of Mixed Modes on Multiple Types of Measurement Error |
| title_fullStr | The Impact of Mixed Modes on Multiple Types of Measurement Error |
| title_full_unstemmed | The Impact of Mixed Modes on Multiple Types of Measurement Error |
| title_short | The Impact of Mixed Modes on Multiple Types of Measurement Error |
| title_sort | impact of mixed modes on multiple types of measurement error |
| topic | multitrait-multierror experimental design measurement error mixed-mode design |
| url | https://ojs.ub.uni-konstanz.de/srm/article/view/7450 |
| work_keys_str_mv | AT alexandrucernat theimpactofmixedmodesonmultipletypesofmeasurementerror AT josephsakshaug theimpactofmixedmodesonmultipletypesofmeasurementerror AT alexandrucernat impactofmixedmodesonmultipletypesofmeasurementerror AT josephsakshaug impactofmixedmodesonmultipletypesofmeasurementerror |