Too many zeros and/or highly skewed? A tutorial on modelling health behaviour as count data with Poisson and negative binomial regression

Health Psychology and Behavioral Medicine, Год журнала: 2021, Номер 9(1), С. 436 - 455

Опубликована: Янв. 1, 2021

Background: Dependent variables in health psychology are often counts, for example, of a behaviour or number engagements with an intervention. These counts can be very strongly skewed, and/or contain large numbers zeros as well extreme outliers. For ‘How many cigarettes do you smoke on average day?’ The modal answer may zero but range from 0 to 40+. same true minutes moderate-to-vigorous physical activity. some people, this near zero, take values someone training marathon. Typical analytical strategies data involve explicit (or implied) transformations (smoker v. non-smoker, log transformations). However, these types ‘counts’ (i.e. non-negative whole numbers) quasi-counts (time is ratio discrete activity could analysed count), and modelled using count distributions – including the Poisson negative binomial distribution (and their zero-inflated hurdle extensions, which alloweven more zeros).Methods: In tutorial paper I demonstrate (in R, Jamovi, SPSS) easy application models data, advantages over alternative ways analysing type two datasets one highly dispersed dependent variable (number views YouTube, another days symptoms were reported month).Results: had best fit overdispersed YouTube. Negative binomial, both good fits symptom over-abundant zeros.Conclusions: cases, provided not just better would lead different conclusions compared poorly fitting traditional regression/linear models.

Язык: Английский

---

Tourism and Mental Health: Foundations, Frameworks, and Futures

Journal of Travel Research, Год журнала: 2022, Номер 62(1), С. 3 - 20

Опубликована: Апрель 18, 2022

Tourism contributes to mental health. We could: recognize, measure, value, and market those contributions; analyze components design products maximize health benefits; use benefits in tourism industry marketing lobbying. If we measure health-research standards, then could also gain commercial opportunities within the sector. Currently, there are widely differing bodies of evidence different subsectors, reflecting historical research emphases. Music, museums, shopping malls have been identified as therapeutic for some. Nature therapies tested extensively, with a higher standard evidence. Adventure has analyzed principally from phenomenological perspective, indicating powerful psychological effects. Many require continuing behavioral change. contribute these therapies, since it includes detailed data on effects program guiding, individual personalities, interests, capabilities, motivations, experiences, emotions, satisfaction.

Язык: Английский

---

The regression trap: why regression analyses are not suitable for selecting determinants to target in behavior change interventions

Health Psychology and Behavioral Medicine, Год журнала: 2023, Номер 11(1)

Опубликована: Окт. 25, 2023

Objective Regression analyses are commonly used for selecting determinants to target in behavior change interventions, but the aim of this article is explain why regression not suitable purpose (i.e. trap).

Язык: Английский

---

The Digital Therapeutics Real World Evidence Framework: An approach for guiding evidence-based DTx design, development, testing, and monitoring (Preprint)

Journal of Medical Internet Research, Год журнала: 2024, Номер 26, С. e49208 - e49208

Опубликована: Янв. 29, 2024

Digital therapeutics (DTx) are a promising way to provide safe, effective, accessible, sustainable, scalable, and equitable approaches advance individual population health. However, developing deploying DTx is inherently complex in that includes multiple interacting components, such as tools support activities like medication adherence, health behavior goal-setting or self-monitoring, algorithms adapt the provision of these according needs may change over time. While myriad frameworks exist for different phases development, no single framework exists guide evidence production across its full life cycle, from initial development long-term use. To fill this gap, we propose real-world (RWE) pragmatic, iterative, milestone-driven approach DTx. The RWE derived 4-phase model used behavioral interventions, but it key adaptations specific unique characteristics ensure highest level fidelity users, also incorporates data (RWD) entire cycle intended any group interested contexts, including those industry, care, public health, academia. Moreover, entities fund research supports agencies regulate might find useful they endeavor improve how DTxcan

Язык: Английский

---

Complexity in psychological self-ratings: implications for research and practice

BMC Medicine, Год журнала: 2020, Номер 18(1)

Опубликована: Окт. 8, 2020

Abstract Background Psychopathology research is changing focus from group-based “disease models” to a personalized approach inspired by complex systems theories. This approach, which has already produced novel and valuable insights into the nature of psychopathology, often relies on repeated self-ratings individual patients. So far, it been unknown whether such self-ratings, presumed observables patient as system, actually display dynamics. We examine this basic assumption psychopathology testing for three markers complexity: memory , presence (time-varying) short- long-range temporal correlations; regime shifts transitions between different dynamic regimes; sensitive dependence initial conditions also known “butterfly effect,” divergence initially similar trajectories. Methods analyzed (1476 time points) single complexity using Bartels rank test, (partial) autocorrelation functions, time-varying autoregression, non-stationarity change point analysis, Sugihara-May algorithm. Results Self-ratings concerning psychological states (e.g., item “I feel down”) exhibited all markers: long-term memory, multiple shifts, conditions. Unexpectedly, physical sensations am hungry”) less dynamics their behavior was more random variables. Conclusions Psychological The in means that we have acknowledge (1) yield pattern data not set (nearly) independent points, (2) humans are “moving targets” whose non-stationary processes including (3) prediction trajectories may be fundamentally impossible. These findings limitation popular statistical series models assumptions violated these markers. conclude mental health should appreciate fundamental aspect adopting methods science. Promising first steps direction, real-time process monitoring, short-term prediction, just-in-time interventions, discussed.

Язык: Английский

---

A scoping review of research on the determinants of adherence to social distancing measures during the COVID-19 pandemic

Health Psychology Review, Год журнала: 2021, Номер 15(3), С. 350 - 370

Опубликована: Май 24, 2021

This scoping review focused on answering key questions about the focus, quality and generalisability of quantitative evidence determinants adherence to social distancing measures in research during first wave COVID-19. The included 84 studies. majority studies were conducted Western Europe USA. Many lacked theoretical input, at risk for bias, few experimental design. most commonly coded domains TDF 'Environmental Context Resources' (388 codes across 76 studies), 'Beliefs Consequences' (34 21 'Emotion' (28 12 'Social Influences' (26 16 studies). least frequently 'Optimism' (not coded), 'Intentions' (coded once), 'Goals' (2 2 'Reinforcement' (3 'Behavioural Regulation' 3 Examining focus identified a lack potentially important such as reinforcement, goal setting self-monitoring. was variable their generalisablity threatened by reliance convenience samples.

Язык: Английский

---

Continuous-Time Modeling of the Bidirectional Relationship Between Incidental Affect and Physical Activity

Annals of Behavioral Medicine, Год журнала: 2022, Номер 56(12), С. 1284 - 1299

Опубликована: Июль 8, 2022

Previous research suggests that there is a bidirectional relationship between incidental affect (i.e., how people feel in day-to-day life) and physical activity behavior. However, many inconsistencies exist the body of work due to lag interval measurements.

Язык: Английский

---

Early Warning Signals in Phase Space: Geometric Resilience Loss Indicators From Multiplex Cumulative Recurrence Networks

Frontiers in Physiology, Год журнала: 2022, Номер 13

Опубликована: Май 4, 2022

The detection of Early Warning Signals (EWS) imminent phase transitions, such as sudden changes in symptom severity could be an important innovation the treatment or prevention disease psychopathology. Recurrence-based analyses are known for their ability to detect differences behavioral modes and order transitions extremely noisy data. As a proof principle, present paper provides example recurrence network based analysis strategy which can implemented clinical setting data from individual is continuously monitored purpose making decisions about diagnosis intervention. Specifically, it demonstrated that measures on geometry space serve transitions. A publicly available multivariate time series analyzed using so-called cumulative Recurrence Networks (cRN), networks with edges weighted by directed towards previously observed points. results compared previous same set, benefits, limitations future directions approach discussed.

Язык: Английский

---

Attractor landscapes: a unifying conceptual model for understanding behaviour change across scales of observation

Health Psychology Review, Год журнала: 2022, Номер 17(4), С. 655 - 672

Опубликована: Ноя. 24, 2022

Models and theories in behaviour change science are not short supply, but they almost exclusively pertain to a particular facet of behaviour, such as automaticity or reasoned action, single scale observation individuals communities. We present highly generalisable conceptual model which is widely used complex systems research from biology physics, an accessible form behavioural scientists. The proposed attractor landscapes can be understand human on different levels, dyads, groups societies. use the tool neglected ideas contemporary science, hysteresis nonlinearity. deepen understanding well-known features (research), including short-livedness intervention effects, problematicity focusing initiation while neglecting maintenance, continuum stage models understood within accommodating framework, concept resilience. also demonstrate potential methods analysis outline avenues for future research.

Язык: Английский

---

Understanding the complexity of individual developmental pathways: A primer on metaphors, models, and methods to study resilience in development

Development and Psychopathology, Год журнала: 2023, Номер 35(5), С. 2186 - 2198

Опубликована: Окт. 10, 2023

Abstract The modern study of resilience in development is conceptually based on a complex adaptive system ontology which many (intersystem) factors are involved the emergence resilient developmental pathways. However, methods and models developed to dynamical systems have not been widely adopted, it has recently noted this may constitute problem moving field forward. In present paper, I argue that an ontological commitment only possible, but highly recommended for development. Such commitment, however, also comes with different causal research methods. first part discuss extent current adheres perspective. second part, introduce conceptual tools help researchers conceptualize causality systems. third discusses idiographic could be used program embraces interaction dominant idiosyncratic nature dynamics conclusion strong warranted, will require radical departure from nomothetic science.

Язык: Английский

---