Mike Brennan

THE BEHAVIOR OF RESPONDENTS, NONRESPONDENTS, AND REFUSERS ACROSS MAIL SURVEYS

Although numerous studies have attempted to measure nonresponse bias by comparing the characteristics of respondents and nonrespondents, most studies have restricted their comparisons to the outcome of a single survey and typically treat nonrespondents as a homogeneous group. However, with regard to mail surveys at least, several studies have noted the importance of distinguishing between nonresponse due to noncompliance and nonresponse due to inaccessibility (Mayer and Pratt 1966; Stinchcombe, Jones, and Sheatsley 1981), raising the possibility that different groups of nonrespondents may have different predispositions toward survey participation. In order to measure a persons predisposition toward survey participation, it is necessary to examine their behavior across surveys. To date, only one published study appears to have done this. Stinchcombe, Jones, and Sheatsley (1981) compared the characteristics and attitudes of early responders, late responders, nonresponders, and refusers to a telephone survey of farmers by examining the responses and participation history of these same farmers to previous mail surveys. They found that the tendency to refuse to participate in surveys was not related to accessibility, or vice versa, and provided evidence that hard-core refusers demonstrated quite different attitudes and behavior from both those easily interviewed and those difficult to contact. Generalizations from Stinchcombe, Jones, and Sheatsleys findings are limited because of the highly specialized sampling frame and topic. But the possibility that respondents, refusers, and other nonrespondents may have different predispositions toward survey participation has important implications for how surveys are conducted and how the effectiveness of survey methods is assessed. The purpose of the

Classification of infant cries using descriptive scales

Attempts to develop rules for identifying infant cry-types have typically relied on acoustic parameters of “distinctive features”. In the present study rules are developed from Semantic Differential (SD) scales. Twenty-four infant cries (six each of pain, birth, hunger, and pleasure) were rated on 50 SD scales by 39 mothers who also completed a cry recognition task. A discriminant function analysis produced classification equations which correctly identified 82% of the 10 criterion items and 64% of the full cry set. Additionally, scales which discriminated the four cry-types were selected using the mean scale ratings of the better recognized cries. Cry-type descriptions were provided by the scale labels. The discriminant function analysis offers a reliable means for classifying cries according to ratings on SD scales, and the SD scales provide cry-type descriptions.

The effect of introductions on telephone survey participation rates

Response rates have been falling across all forms of survey research for decades (see for example Baim 1991; Meier 1991; Smith 1995; Groves & Couper 1998; de Leeuw & de Heer 2002; Bednall & Shaw 2003; CMOR 2003), and this trend seems likely to continue. Thus, given that all survey research depends on respondent cooperation to ensure valid and unbiased results, it is imperative that survey researchers develop and adopt techniques that can be used to increase rates of survey participation. This paper reports on four techniques that may be used in telephone surveys.

The Effect of Product Sampling and Couponing on Purchase Behaviour: Some Empirical Evidence

This article reports the results of three experiments that assessed the performance of mail-drop product samples, discount coupons and cashbacks among non-users of three consumer products: liquid detergent, instant coffee and childrens toothpaste. For each product, households were assigned to one of four treatment groups: sample plus discount coupon or cashback; sample only; discount coupon or cashback only, and a control group (no promotion). Treatments which included samples achieved significantly higher purchase rates than treatments which did not, while including discount coupons with samples produced slightly higher purchase rates than samples delivered alone. Cashback coupons were as effective as product samples in producing significantly higher purchase rates than the no promotion control. These results suggest both product sampling and cashbacks promote incremental sales, but that discount coupons may only subsidize purchases that would otherwise be made at full retail prices.

Using the Juster Scale to Estimate the Demand-Price Relationship

Abstract The Juster scale was used to estimate the price-demand relationship for two branded products, Coca-Cola in cans, and Campbells Red and White Label soup. For each product predictions were obtained of the number of items that would be purchased, during the following four weeks, at each of two hypothetical prices, one above and one below the current retail price. Predictions were obtained using two variations on the Juster procedure: the multiple question method and the constant sum method. Actual numbers purchased were obtained in follow-up interviews. Actual purchases of Coca-Cola were almost exactly at the predicted level, but purchases of the soup brand were less than a third of those predicted. This is the first reported use of the Juster Scale to predict purchases of brands as opposed to products, or to estimate price-demand relationships.

Perceptual Dimensions of Infant Cry Signals: A Semantic Differential Analysis:

In the present study the Semantic Differential technique was used to examine perceived characteristics of infant cry sounds. 24 cries (6 each of pain, hunger, birth, and “pleasure”) were rated by 39 mothers on 50 scales. A factor analysis of the scale ratings uncovered three main factors, labelled Affect, Potency, and Evaluation, respectively. 5 “factorially-pure” scales were selected to represent each factor. A hierarchical cluster analysis of the mean factor-scale ratings uncovered 3 major cry clusters which essentially represented 3 cry-type groups, hunger, pleasure, and pain/birth. A close correspondence was found between this set of clusters and another set generated independently from cry-recognition data, indicating that the semantic differential factor-scales effectively discriminated perceptually distinct cries. This procedure adequately separates the physical (acoustic) properties of signals from their listener-perceived (auditory) ones. This distinction is not possible with studies which depend upon multivariate techniques and casts doubt on their findings.

14.5 Development in Marketing Research: Analytical Techniques: New Applications

The Juster Scale is an eleven-point purchase probability scale that has provided a reasonably accurate method for predicting purchase rates (the proportion of people who will buy). This study examined the accuracy of three methods for using the Juster Scale to predict purchase levels, that is, the quantity of product people will buy. While two methods produced reasonably accurate predictions for most items, they can only be used in a face-to-face situation. A simplified procedure produced even more accurate results, and may be suitable for telephone interviews.

Caregivers’ Reactions To Infant Cries and Faces

Semantic Differential scales were used to examine caregivers’ reactions to video segments of the faces and cries of two full‐term and two premature infants. Some segments comprised cry sounds only; some comprised cry faces only; and others comprised cry + face combinations where the cry and face were either from a single infant or from different infants, one premature and the other full‐term. The main findings were: (i) while the features of the cries and faces of the premature infants were rated more negatively than those of the full‐term infants, they did not evoke more negative affective reactions from the viewers, (ii) while the negative characteristics of one premature infant were primarily associated with its cries, the negative ratings of the other premature infant were primarily associated with its face, and (iii) the ratings of the faces and cries separately were not predictive of the ratings of the cry + face combinations. These findings are discussed in relation to research suggesting links bet...

Profiles Obtained from Clustering Behaviours in Situations Rated on Infant Temperament Dimensions

Abstract When working with parents trying to cope with an exceptional child in a family setting, it is often difficult to meaningfully illustrate their views of the childs current behaviours and/or changes brought about by deliberate interventions. One approach outlined in the present paper involves obtaining from participants their ratings on a number of bi‐polar “temperament” dimensions of behaviours which a child may display in different situations. These data are then subjected to a cluster analysis which displays correspondences graphically, in a form which can be discussed directly. An example drawn from the authors’ involvement with parents coping with crying babies is provided.