Exploring Best–Worst Scaling for Communication Sciences and Disorders

Purpose The primary purpose of this study is to introduce the method of best–worst scaling (BWS) and explore how it can be applied to the field of communication sciences and disorders. Method Two hundred thirty-six participants (156 students and 80 instructors) responded to a survey using BWS to identify the most and least preferred features of a speech-language pathology graduate program. Results were analyzed with Sawtooth Software and best–worst values were transformed into a standardized ratio scale. Results The number one attribute that students were most concerned with when choosing a graduate program was clinical placements followed by financial aid/scholarships/graduate assistant positions, highly ranked/reputation of program, updated resources available in clinic, reputation of professors, elective coursework available in an interest area, proximity to home, research opportunities, completion of an undergraduate degree at the same institution, and, finally, the availability of online courses. Five of the 10 factors were statistically different from the adjacent factor as evidenced by nonoverlapping confidence intervals. The faculty accurately predicted the top three and bottom two attributes but distinguished less between the factors as evidenced by overlapping confidence intervals in eight of the adjacent pairs. Conclusions BWS is able to provide nuanced information relative to the strength of the preference over other types of importance measures. It has many applications in the field of communication sciences and disorders, and this pilot study revealed one way it can be utilized. Knowing what students most desire in a graduate program allows universities to tailor their marketing to highlight the attributes of highest importance to future applicants.

A146 MANAGEMENT OF ANTITHROMBOTIC THERAPY AFTER GASTROINTESTINAL BLEEDING: A MIXED METHODS STUDY OF HEALTHCARE PROVIDERS

Background Oral anticoagulants (OAC) are permanently discontinued in up to 50% of patients after gastrointestinal (GI) bleeding despite ongoing thrombotic risk and evidence of benefit to restarting. The reasons for permanent discontinuation of OAC are unclear, but likely include concerns about re-bleeding and a lack of high-quality evidence. There are no studies evaluating healthcare provider values and preferences following OAC-related GI bleeding and their influence on decision-making about whether and when to resume OACs. Aims We aimed to (i) identify key factors (attributes) that influence healthcare provider decision-making regarding resumption of OAC after GI bleeding, (ii) determine the relative importance of these attributes, and (iii) to identify preference groups. Methods We conducted focus group discussions (FGD) with healthcare providers involved in the care of patients with OAC-related GI bleeding. Participants ranked the relative importance of the attributes identified during the FGD through a dot voting exercise. Transcripts were reviewed and themes (attributes) were identified. Results of the FGD informed a discrete choice experiments survey developed and analyzed using the Sawtooth software platform (Sawtooth Software, USA). The survey was pilot tested and modified by iterative feedback. A sample choice task is shown in Image 1. Hierarchical Bayes analysis was used to estimate preference coefficients (utilities) for each attribute. Latent class analysis was used to identify preference groups. Results There were 4 FGD involving 29 participants. The most important attributes identified and included in the survey were thrombosis risk, indication for OAC, index bleed severity, re-bleeding risk, and patient characteristics. There were 130 survey respondents practicing in hematology (n=68), internal medicine (n=30), gastroenterology (n=7), cardiology (n=4), family medicine (n=3), and others (n=18). The mean age was 45 years (±11) and 51% were male. Thrombosis risk and re-bleeding risk equally had the highest utility followed by index bleed severity, patient characteristics, and indication for OAC. Two preference groups were identified. The dominant preference group (78% of respondents) placed the highest utility on thrombosis risk and re-bleeding risk, while a minority (22% of respondents) placed the highest utility on index bleed severity. Conclusions Thrombosis risk and re-bleeding risk are equally the most important factors influencing OAC resumption following OAC-related GI bleeding. The severity of the index bleed is the most important factor in decision-making for a minority segment of healthcare providers. Further research on the dose, type and timing of OAC resumption is needed to determine the optimal balance between thrombosis and re-bleeding. Funding Agencies Heart and Stroke Foundation

All aboard the corporate socially and environmentally responsible cruise ship

Purpose The purpose of the study was to examine the relative importance of corporate social and environmental responsibility (CSER) in comparison to standard, price, duration, destination, brand and disruption using choice-based conjoint analysis (CBC). Design/methodology/approach CBC was used as the data collection survey technique, and counts analysis for preference and hierarchical Bayes estimation (HB) for importance levels data analysis methods, from Sawtooth Software Inc. Findings Results show that 2:1 Royal Caribbean Cruise Line cruise consumers prefer companies with CSER policies and practices. However, their actual product choice selection of cruise package attributes revealed that consumers overall placed less importance on CSER when choosing cruises. Experienced consumers were more brand image-conscious than those new to cruising, and consumers who were less price-sensitive were most willing to choose companies with CSER policies and practices. Research limitations/implications The information provided is specifically on “what” cruise consumer preferences and importance attributes are but does not explicitly explain “why” the respondents made the choices they did. This was at the time a limitation of the software used to conduct the study. Practical implications The Conjoint Analysis CBC Sawtooth Software pre-2014 version choice simulators do not facilitate questions that provide answers as to “why” respondents make the choices they do in the market simulations. Social implications The knowledge contribution is of value to both academia and industry, as the quantitative statistical data on the cruise consumers’ choice preferences are of value in understanding and identifying solutions/approaches towards “opening the bottleneck” that exists between private sector sustainable development practices and consumer lifestyle changes. Originality/value This was the first time that CBC/HB was applied within academia to examine the cruise consumers’ choice preferences in a UK context and also the first time that CSER was applied as a direct variable in a cruise package to determine the preference and important values of a brand in a consumer behaviour decision-making context.

Une comparaison empirique de la validité prédictive de la méthode de composition, de l'analyse conjointe et de l'analyse conjointe hybride

Cette recherche évalue l'efficacité de différentes méthodes de modélisation des préférences des consommateurs au niveau individuel : la méthode de composition (phase I et II d'ACA4 de Sawtooth Software), l'analyse conjointe traditionnelle (Conjoint Analyzer de Bretton Clark; profil complet) et l'analyse conjointe hybride (ACA4 de Sawtooth Software). Les conclusions principales sont les suivantes : dans le cas où le nombre d'attributs est inférieur à six, la validité prédictive de l'analyse conjointe hybride et de l'analyse conjointe traditionnelle (profil complet) est supérieure à celle de la méthode de composition. En revanche, la validité prédictive de l'analyse conjointe hybride est inférieure à celle de l'analyse conjointe traditionnelle (profil complet).