The frequent use of CAWI over the last few years has determined two needs: firstly, the evaluation of factors that determine the success of the survey in terms of overall participation; secondly, the specification of a model that enables ALMALAUREA to have an ex ante forecast of the final response rate. The analysis presented in this paper will focus on approximately ten CAWI surveys conducted during the last few years. They mostly regard the employment opportunities of both pre- and post-reform graduates one or more years on from graduation. Sometimes ad hoc surveys have also been carried out to explore more specific topics, e.g., graduates’ opinion on interculturalism. These surveys, which lasted about 3-6 weeks each, have produced response rates ranging from 31 to 49 percent. Although the surveys have different characteristics – in terms of topics, question texts, duration of data collection, day of the week and time of day when data collection starts, number of reminders sent, and so on – a preliminary analysis revealed a common trend: the utmost participation of graduates is observed during the first few days immediately after the beginning of the survey itself; afterwards, there is a gradual reduction in participation. Moreover, the contribution given by reminders is valuable and immediate: each time a reminder has been sent there was an increase in the number of questionnaires answered, but they have a limited effect over the course of time. The analysis of response rate trends reveals that the final response rate is particularly connected to the participation rate recorded during the initial days of data collection. The first objective of this paper is, therefore, to find a function that calculates the overall response rate achieved during an online survey on the basis of both contextual factors (i.e., the elements that are peculiar to the survey) and individual factors (i.e., the characteristics of the population). But what are the decisive factors for the success of an online survey? Thanks to the wide range of information available to ALMALAUREA – deriving both from administrative sources and from graduates CVs – a preliminary descriptive analysis highlights the fact that respondents of online surveys comprise a self-selected sample. The second objective of this paper is to use a logistic regression model in order to identify the variables that most influence graduates’ probability of participating in an online survey and calculate their net effect. In this way it will be possible to maximise the success of the survey and to keep self-selection under check by intervening on the most relevant variables.

Modelling Online Survey Participation among Italian University Graduates

GASPERONI, Giancarlo
2009

Abstract

The frequent use of CAWI over the last few years has determined two needs: firstly, the evaluation of factors that determine the success of the survey in terms of overall participation; secondly, the specification of a model that enables ALMALAUREA to have an ex ante forecast of the final response rate. The analysis presented in this paper will focus on approximately ten CAWI surveys conducted during the last few years. They mostly regard the employment opportunities of both pre- and post-reform graduates one or more years on from graduation. Sometimes ad hoc surveys have also been carried out to explore more specific topics, e.g., graduates’ opinion on interculturalism. These surveys, which lasted about 3-6 weeks each, have produced response rates ranging from 31 to 49 percent. Although the surveys have different characteristics – in terms of topics, question texts, duration of data collection, day of the week and time of day when data collection starts, number of reminders sent, and so on – a preliminary analysis revealed a common trend: the utmost participation of graduates is observed during the first few days immediately after the beginning of the survey itself; afterwards, there is a gradual reduction in participation. Moreover, the contribution given by reminders is valuable and immediate: each time a reminder has been sent there was an increase in the number of questionnaires answered, but they have a limited effect over the course of time. The analysis of response rate trends reveals that the final response rate is particularly connected to the participation rate recorded during the initial days of data collection. The first objective of this paper is, therefore, to find a function that calculates the overall response rate achieved during an online survey on the basis of both contextual factors (i.e., the elements that are peculiar to the survey) and individual factors (i.e., the characteristics of the population). But what are the decisive factors for the success of an online survey? Thanks to the wide range of information available to ALMALAUREA – deriving both from administrative sources and from graduates CVs – a preliminary descriptive analysis highlights the fact that respondents of online surveys comprise a self-selected sample. The second objective of this paper is to use a logistic regression model in order to identify the variables that most influence graduates’ probability of participating in an online survey and calculate their net effect. In this way it will be possible to maximise the success of the survey and to keep self-selection under check by intervening on the most relevant variables.
2009
ITACOSM09 - First ITAlian COnference on Survey Methodology 2009 - Book of Abstracts
264
266
Cimini C.; Girotti C.; Gasperoni G.
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/100955
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact