Survey Non-Response

To address the problem on nonresponse in sample surveys, Hansen and Hurwitz
(1946) have developed an estimator and planned the procedure of sub-sampling from
non-respondents and re-contact them to obtain the information using different data
collection method and assume that no non response is observed in this subsample. But in
recent sampling it is not practicable, some non-response may occur in the subsample. In
this paper, we developed estimators when non-response occurs in subsample(s). The
expressions for variance of the proposed estimators have been derived.

Identifying ways to efficiently maximize the response rate to surveys is important in survey-based research. However, evidence on the response rate effect of donation incentives and especially altruistic and egotistic text appeal interventions is sparse and ambiguous. Via a randomized survey experiment among 6,162 members of an online survey panel, this article shows how low-cost incentives and cost-free text appeal interventions may affect the survey response rate in online panels. The experimental treatments comprise (a) a cash prize lottery incentive, (b) two donation incentives that promise a monetary donation to a good cause in return for survey response, (c) an egotistic text appeal, and (d) an altruistic text appeal. Relative to a control group, we find higher response rates among recipients of the egotistic text appeal and the lottery incentive. Donation incentives yield lower response rates.

El trabajo se divide en cuatro capı́tulos autocontenidos. En el primero se presenta el
trabajo de Moreno titulado “Clustering robusto basado en modelos”. En este documento
el autor analiza y compara distintas técnicas de clustering basadas en modelos proba-
bilı́sticos con el objetivo de encontrar el procedimiento más eficiente y estable entre el
conjunto analizado, considerando diferentes tipologı́as de datos, en presencia de outliers.
Se prueba este enfoque en la conformación de grupos para un conjunto de inmuebles de
Montevideo y Canelones a partir del precio de ellos, los metros cuadrados construidos y
del terreno en el que se aloja el inmueble.
En el siguiente capı́tulo, se presenta el trabajo de Coimbra y Antı́a titulado “Trata-
miento de la no respuesta en encuestas de panel en el caso de poblaciones finitas”, cuyo
objetivo es abordar una forma de tratamiento de la no respuesta en un tipo especı́fico de
encuestas por muestreo, las encuestas de panel.
En el capı́tulo tres, en una lı́nea próxima al trabajo de Moreno, se presenta el trabajo
de Massa cuyo tı́tulo es “Efecto de valores faltantes en estudios longitudinales en adultos
mayores”. Es ente trabajo se aborda la forma de solucionar el problema del abandono de
los sujetos incluidos en estudios denominados Mini Mental State Examination (MMSE),
se centra en el uso de modelos conjuntos donde el tiempo de sobrevida de los sujetos y
los resultados del MMSE están relacionados.
Finalmente, en el último capı́tulo, se presenta el trabajo de Vernazza en el que se
analizan las propiedades psicométricas de un instrumento de medición de la satisfacción
estudiantil en nuestra Facultad de Ciencias Económicas y de Administración.

In this paper effect of measurement error and non response error is examined on estimation of unknown population mean of study variable. We have obtained the expression of the MSE (mean square error) of the proposed estimator up to first order of approximation. We have shown theoretically and empirically that the proposed estimator performs better than other estimators considered in this article. For empirically study we have used for different data sets. RESUMEN En este paper es examinado el efecto de error de medición de respuesta en la estimación de la media desconocida de la variable de estudio. Hemos obtenido la expresión del MSE (error cuadrático medio) del estimador propuesto hasta el primer orden de aproximación. Hemos mostrado teoréticamente y empíricamente que el propuesto estimador se comparta mejor que los otros considerados en este artículo. Para los estudios empíricos usamos diferentes data.

A lot of work has been done in two-phase sampling for estimating the population mean of study variable while considering the non-response at second phase see e.g. Khare and Srivastava (1993, 1995), Tabasum and Khan (2004), Singh and Kumar (2008) and Singh et al. (2010). Most of authors have used information of single auxiliary variable while estimating the mean of study variable, whereas in practical situation we may require/have auxiliary information on multiple characters. Khair and Sinha (2009, 2011) have used multi-auxiliary characters for estimating the population mean in the presence of nonresponse in the case of simple random sampling. In two-phase sampling, when auxiliary information is obtained at first phase sample that is relatively large sample than second phase, the nonresponse rate will be high as compared to second phase. In this paper we have proposed the class of generalized estimators for estimating population mean of study variable under two phase sampling scheme using multi-auxiliary variable(s) in the presence of non-response at first phase. The information on all auxiliary variable(s) is not known for the population. The bias and mean square error have also been derived. Empirical study has been conducted for comparing the efficiency of proposed estimators.

Резюме. Парадані мають потенційну цінність для вимірювання та поліпшення якості даних, а також для скорочення витрат у процесі збору даних, саме тому вони представляють інтерес не тільки для дослідницьких організацій, а також для замовників досліджень, аналітиків і експертів з методології кількісних досліджень. У роботі розглядаються можливості використання параданих для зменшення рівня невідповідей у емпіричних соціологічних дослідженнях на прикладі дослідження ESS, аналізуються та класифікуються причини відмови від проходження опитування та шляхи їх уникання.
Abstract. Paradata have potential value to measure and improve the quality of data, as well as to reduce costs in the data collection process, so they are of interest not only for research organizations but for customers, analysts and experts on the methodology of quantitative research. In the paper describes the possibilities to reduce with paradata the non-response level in empirical sociological research case study of ESS, analyzed and classified the reasons for refusal of questioning and ways to avoid them.

The purpose of any statistical analysis is to uncover the trend using real situations and provide an accurate model for future policies. In order to fulfill this aim, it is important to collect and use the data carefully and minimize the possibility of various types of errors. Nonresponse (NR) and measurement error (ME) are the two major types of nonsampling error that occur in almost every survey. We propose an estimator of the population mean and study its efficacy in the combined and separate effects of NR and ME. Some existing estimators are also considered for comparison. A simulation study was performed to see its efficiency numerically.

Osnovni cilj rada je utvrđivanje efekta klasne pripadnosti, strukturnih činilaca (pola, starosti, mesta stanovanja) i političke participacije na učestalost i distribuciju odgovora „ne znam“ ili „bez odgovora“ na nizu empirijskih iskaza, dobijenih putem anketnog istraživanja stanovništva Srbije iz 2018. godine, koji mere stavove građana o različitim društvenim pitanjima. Teorijske osnove analize izvedene su iz komplementarne upotrebe dva stanovišta: prvo su teorije o postpolitici i postdemokratiji Žaka Ransijera, Kolina Krauča i drugih autora, koje ukazuju na depolitizaciju i dedemokratizaciju savremenih (neoliberalnih) društava, dok je drugo Burdijeovo stanovište o opštim uslovima proizvodnje političkog mišljenja u zapadnim klasnim demokratijama.

We will study the use of RSS as alternative to SRSWR design when Saha´s RR is implemented. Estimators of the mean are derived under 3 alternative ranking criteria. The expected variances are obtained and used for comparing their accuracy. The RSS alternatives are superior in terms of having a smaller sampling error than the SRSWR design..