Selection of influential variables in ordinal data with preponderance of zeros

Presence of excess zero in ordinal data is pervasive in areas like medical and social sciences. Unfortunately, analysis of such kind of data has so far hardly been looked into, perhaps for the reason that the underlying model that fits such data, is not a generalized linear model. Obviously some methodological developments and intensive computations are required. The current investigation is concerned with the selection of variables in such models. In many occasions where the number of predictors is quite large and some of them are not useful, the maximum likelihood approach is not the automatic choice. As, apart from the messy calculations involved, this approach fails to provide efficient estimates of the underlying parameters. The proposed penalized approach includes ?₁ penalty (LASSO) and the mixture of ?₁ and ?₂ penalties (elastic net). We propose a coordinate descent algorithm to fit a wide class of ordinal regression models and select useful variables appearing in both the ordinal regression and the logistic regression based mixing component. A rigorous discussion on the selection of predictors has been made through a simulation study. The proposed method is illustrated by analyzing the severity of driver injury from Michigan upper peninsula road accidents.     

Projecting total costs and health consequences of increasing mt-sDNA utilization for colorectal cancer screening from the payer and integrated delivery network perspectives

Abstract

Aims: To evaluate total costs and health consequences of a colorectal cancer (CRC) screening program with colonoscopy, fecal immunochemical tests (FIT), and expanded use of multitarget stool DNA (mt-sDNA) from the perspectives of Integrated Delivery Networks (IDNs) and payers in the United States.

Materials and methods: We developed a budget impact and cost-consequence model that simulates CRC screening for eligible 50- to 75-year-old adults. A status quo scenario and an increased mt-sDNA scenario were modeled. The status quo includes the current screening mix of colonoscopy (83%), FIT (11%), and mt-sDNA (6%) modalities. The increased mt-sDNA scenario increases mt-sDNA utilization to 28% over 10 years. Costs for both the IDN and the payer perspectives incorporated diagnostic and surveillance colonoscopies, adverse events (AEs), and CRC treatment. The IDN perspective included screening program costs, composed of direct nonmedical (e.g. patient navigation) and indirect (e.g. administration) costs. It was assumed that IDNs do not incur the costs for stool-based screening tests or bowel preparation for colonoscopies.

Results: In a population of one million covered lives, the 10-year incremental cost savings incurred by increasing mt-sDNA utilization was $16.2 M for the IDN and $3.3 M for the payer. The incremental savings per-person-per-month were $0.14 and $0.03 for the IDN and payer, respectively. For both perspectives, increased diagnostic colonoscopy costs were offset by reductions in screening colonoscopies, surveillance colonoscopies, and AEs. Extending screening eligibility to 45- to 75-year-olds slightly decreased the overall cost savings.

Limitations: The natural history of CRC was not simulated; however, many of the utilized parameters were extracted from highly vetted natural history models or published literature. Direct nonmedical and indirect costs for CRC screening programs are applied on a per-person-per modality basis, whereas in reality some of these costs may be fixed.

Conclusions: Increased mt-sDNA utilization leads to fewer colonoscopies, less AEs, and lower overall costs for both IDNs and payers, reducing overall screening program costs and increasing the number of cancers detected while maintaining screening adherence rates over 10 years.     

The mitigating effect of matching regulatory focus with arousal-inducing stimuli in service failure situations

Service failures are pivotal touchpoints that can reduce customer satisfaction, encourage negative word-of-mouth, and ultimately impact a firm's market share. We advance a novel perspective that after a service failure occurs, matching incidental arousal-inducing stimuli to one's regulatory orientation can make the negative experience stemming from the service failure less deleterious. In three experiments (two stock out scenarios and one involving a rude salesperson), after a service failure, promotion-focused and prevention-focused individuals were exposed to high versus low arousal-inducing stimuli. Three approaches available to retailers were used to manipulate arousal levels: background pictures (Study 1), colors (Study 2), and music (Study 3). When high (low) incidental arousal-inducing stimuli were presented to those with a promotion (prevention) focus, this raised satisfaction, loyalty, and referral for brands compared to when promotion (prevention)-focused individuals were exposed to low (high) arousal-inducing stimuli. Changes in self-rated arousal and affect valence levels (arousal and valence levels were measured after the service failure and then after exposure to the incidental arousal-inducing stimuli) mediated the effect on these consumer behaviors. These insights extend theory by considering the combined effect of regulatory focus and affect. They also have practical relevance.     

Swine flu and hype: a systematic review of media dramatization of the H1N1 influenza pandemic

Highly disconcerting at the time, in retrospective, the 2009 H1N1 influenza pandemic looks like much ado about nothing. As a consequence, many accused the media of having created an artificial hype or hysteria around the new virus, thus contributing to unwarranted public fear. The current paper set out to examine the validity of such accusations. We integrated empirical findings on whether the media dramatized H1N1 on a global scale through systematically reviewing prior content-analytic studies. We developed a coding scheme specifying three indicators of dramatized media coverage that – together – inform about how mass media coverage about H1N1 may amplify risk perceptions in the public: (a) the volume of media coverage, (b) the media content presented, particularly an overemphasis of threat while neglecting measures of self-protection and (c) the tone of coverage. Results show that media attention was immense, that news content stressed threat over precautionary measures, while the pattern of coverage tonality remained nebulous due to conflicting findings. The present review also revealed a critical gap in existing knowledge about the tone of media coverage on H1N1, and discusses implications for future research on dramatization of public health risks by the media.     

Do Laws Influence the Cost of Real Estate Brokerage Services? A State Fixed Effects Approach

A FTC‐DOJ study argues that state laws and regulations may inhibit the unbundling of real estate brokerage services in response to new technology. Our data show that 18 states have changed laws in ways that promote unbundling since 2000. We model brokerage costs as measured by number of agents in a state‐level annual panel vector autoregressive framework, a novel way of analyzing wasteful competition. Our findings support a positive relationship between brokerage costs and lagged house price and transactions. We find that change in full‐service brokers responds negatively (by well over two percentage points per year) to legal changes facilitating unbundling.     

Accounting-based versus market-based cross-sectional models of CDS spreads

Models of financial distress rely primarily on accounting-based information (e.g. [Altman, E., 1968. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. Journal of Finance 23, 589–609; Ohlson, J., 1980. Financial ratios and the probabilistic prediction of bankruptcy. Journal of Accounting Research 19, 109–131]) or market-based information (e.g. [Merton, R.C., 1974. On the pricing of corporate debt: The risk structure of interest rates. Journal of Finance 29, 449–470]). In this paper, we provide evidence on the relative performance of these two classes of models. Using a sample of 2860 quarterly CDS spreads we find that a model of distress using accounting metrics performs comparably to market-based structural models of default. Moreover, a model using both sources of information performs better than either of the two models. Overall, our results suggest that both sources of information (accounting- and market-based) are complementary in pricing distress.     

How product–cause fit and donation quantifier interact in cause-related marketing (CRM) settings: evidence of the cue congruency effect

We are the first to examine the joint impact of product–cause fit and donation quantifier in the cause-related marketing (CRM) domain. We show that these two CRM cues interact in a unique manner, reflecting the cue congruency effect. Specifically, congruent combinations of these two cues result in high purchase intentions when the cues individually have positive effects. In all other cases, however, purchase intentions are low. Furthermore, we identify moderators of the above cue congruency effect. In Study 1, we show that the cue congruency effect is moderated by product-type, evidencing only in more hedonic product contexts. In Study 2, we show that the above cue congruency effect is moderated by purchase-type, evidencing in planned purchase contexts, but reversing in impulse purchase contexts. We discuss the process mechanism driving these effects, specify the contribution of this research for CRM, cue congruency and impulse purchases, and outline implications for practice.