Mixed Effects Models

This paper examines the usefulness of customer satisfaction to analysts when preparing their earnings forecasts. We draw on theory in marketing to predict how customer satisfaction should be associated with earnings forecasts and forecast errors. We assembled a dataset of companies studied in the American Customer Satisfaction Index (University of Michigan), which also appear on the Institutional Brokers Estimate System (I/B/E/S) files. By combining these sources, we were able to analyze the forecast errors of 1,875 analysts following 90 companies yielding 8,034 year-firm-analyst observations. We control for factors known to influence the earnings forecasts, such as firm profitability and risk, as well as potential unobservable factors using a Mixed-effects regression. We find that customer satisfaction has a negative association with the analysts' forecast errors because it allows analysts forecasts to be closer to the business reality. The influence of customer satisfaction v...

This paper considers the modelling choices available to researchers using multilevel data, including longitudinal data of various types. Specifically, we consider fixed effects (FE) and random effects (RE) models, including the within-between RE model, often misleadingly termed the ‘hybrid’ model. We argue that the latter is unambiguously a RE model, and that it is the most general of the three models, and as such a sensible starting point, given its flexibility to incorporate the positive aspects of both FE and RE models, and its ability to allow extensions (such as random slopes) that are often important. We present simulations that reveal the extent to which these models cope with mis-specification, finding that failing to include random slopes (e.g. in a FE or standard RE model) can lead to anti-conservative standard errors, and that mis-specifying non-Normal random effects as Normally distributed can introduce some small biases to variance and random effect estimates, but not fixed-part parameter estimates. We conclude with advice for applied researchers, and present a glossary, that gives clear definitions to terms that are confusing or have more than one meaning. Overall, we hope the paper gives practical advice to researchers in many different social science disciplines and beyond, looking to understand and use multilevel and longitudinal data.

In the Philippines, very high fishing pressure coincides with the globally greatest number of shorefish species, yet no long-term fisheries data are available to explore species-level changes that may have occurred widely in the most species rich and vulnerable marine ecosystem, namely coral reefs. Through 2655 face-to-face interviews conducted between August 2012 and July 2014, we used fishers' recall of past catch rates of reef-associated finfish to infer species disappearances from catches in five marine key biodiversity areas (Lanuza Bay, Danajon Bank, Verde Island Passage, Polillo Islands and Honda Bay). We modeled temporal trends in perceived catch per unit effort (CPUE) based on fishers' reports of typical good days' catches using Generalized Linear Mixed Modelling. Fifty-nine different finfish disappeared from catches between the 1950s and 2014; 42 fish were identified to species level, two to genus, seven to family and eight to local name only. Five species occurring at all sites with the greatest number of fishers reporting zero catches were the green bumphead parrotfish (Bolbometopon muricatum), humphead wrasse (Cheilinus undulatus), African pompano (Alectis ciliaris), giant grouper (Epinephelus lanceolatus) and mangrove red snapper (Lutjanus argentimaculatus). Between the 1950s and 2014, the mean perceived CPUE of bumphead parrotfish declined by 88%, that of humphead wrasse by 82%, African pompano by 66%, giant grouper by 74% and mangrove red snap-per by 64%. These declines were mainly associated with excess and uncontrolled fishing, fish life-history traits like maximum body size and socioeconomic factors like access to market infrastructure and services, and overpopulation. The fishers' knowledge is indicative of extirpations where evidence for these losses was otherwise lacking. Our models provide information as basis for area-based conservation and regional resource