Greater thalamic response to risky reward versus loss feedback in PSUs is consistent with research demon strating that thalamic BOLD signals are linked to relapse in cocaine-dependent individuals . Thalamus acts as a relay center for the brain by sending sensory information to insula for further interoceptive processing ; hypo activation to loss may reflect differences in relay and integration of information during decision making. With respect to baseline characteristics, DSUs endorsed higher baseline levels of state depression than PSUs, which may have affected RGT performance given that individuals with depression tend to be risk averse . However, given that mean scores for DSUs are substantially below the Beck Depression Inventory threshold for clinical depression [in nonclinical populations, scores above 20 indicate depression ; it is unlikely that DSUs performed in a manner consistent with samples with depression.Across OSUs, lower frontal, temporal, parietal, insular, and thalamic BOLD signals during risky decisions compared with safe decisions predicted greater future marijuana use . These regions are considered important for executive functions such as inhibitory control, working memory, and attention as well as for being relay centers for integrating information critical for decision making . Therefore, blunted responses in these regions while making choices between risky and safe options may predispose young adults to repeatedly choose marijuana consumption despite potential negative consequences . While cumulative marijuana uses between study visits was related to baseline BOLD patterns, lack of relationship between cumulative interim stimulant use and baseline BOLD signal suggests that while adose–response effect may exist between brain activation and marijuana use,outdoor vertical farming the relationship between brain activation and stimulant use may be better defined through a categorical perspective that includes accompanying clinical sympto mology.
Although PSUs and DSUs used marijuana at significantly high rates , groups did not differ categorically in marijuana abuse/dependence frequency. In contrast, stimulant use in and of itself might not be related to brain differences unless it is accompanied by clinical problems, suggesting that a categorical perspective is a more useful way to conceptualize differences.This study has several unique strengths, including its longitudinal design, use of a model previously applied to chronic stimulant users, and assessment of substance use from both categorical and dimensional perspectives . However, this study is limited by our sample’s significant co-use of marijuana and the categorical criteria that prioritized differences as a function of SUD over marijuana use disorder given that PSUs and DSUs did not differ on baseline/interim marijuana use. In addition, although SUD has been associated with greater incidence of psychiatric illness , lack of clinical symptom measures collected at follow-up hinders our ability to determine whether mental health symptoms affected interim substance use. We are also limited by an inability to evaluate the RGT from a trial by-trial perspective to determine whether BOLD response patterns translate into future behavior or are affected by the preceding trial; due to the limited number of separate 40 and 80 trials, it would not be possible to obtain sufficient statistical power to conduct such an analysis.Project work in a statistics class is a valuable tool for giving students a context to the data they are using and a motivation for learning statistical reasoning . Projects also give the students an appreciation of the practical issues involved in carrying out experiments and collecting data, an outcome encouraged by Higgins . However there are also practical issues in implementing student projects in a curriculum, particular in large classes. For example, our biomedical students are very keen to use friends and family as subjects in their statistics experiments, raising ethical concerns, while other students want to use equipment and resources that are beyond the scope of an introductory statistics course. The result has been that the students often end up doing simplistic experiments which may actually reinforce a trivial view of the role of data analysis in science.
Such environments are particularly useful in helping students understand issues in experimental design by giving them more complex settings than they would have access to in a real experiment, as in the industrial process and greenhouse simulations of Darius et al. or the virtual vaccination trial of Duchateau et al. . In this paper we present an online environment, the Island, where students can conduct studies involving virtual human subjects. Again the first aim of the environment is to engage the students in thinking about the design of the experiment given a statistical question of interest. The second aim is to provide them with data from their design that they can then use in learning statistical methodology. The Island involves two main simulations running at different timescales. In Section 2 we give an overview of the historical simulation that forms the basis of the Island population while in Section 3 we describe the simulation that runs in the present to allow students to conduct experimental studies. In both sections we will outline the design of the corresponding simulations and their role in supporting learning. We emphasize this in Section 4 with four examples of student engagement with the Island. An initial motivation for this system was the need for a virtual environment where students could collect data for addressing questions in epidemiology. While it is easy to generate some simulated data for a randomized clinical trial, for example, our belief is that thinking about issues in epidemiology requires access to a population that can be studied more deeply. Important requirements include the ability to consider the ancestors of particular individuals, to determine whether diseases have a genetic component, and to explore geographic relationships between individuals in order to look for infective characteristics. For these reasons our aim was to simulate a population over time in a spatial context. The Islanders live in 39 villages that range in population from just 26 to 2,292 and provide environmental effects in the simulation of the Islanders. Figure 1 shows samples of the Islander images that are included in the interface, created using the approach discussed in Bulmer and Engstrom . The Island map and other features of the interface can be seen online using the link in Section 1.1. Within each village the Islanders then live within houses. Many live in families while others live as couples or alone.
Again there are environmental effects tied to houses,vertical farming benefits such as an increased chance of taking up smoking if other people in the house already smoke. The location of the house is also important since various diseases are linked to particular parts of the Island and transmissible diseases are spread based on distance between individuals. Sampling an Islander at random is deliberately a difficult task. Each village does have a hall with records of births, deaths and marriages but Islanders do move around during their lives. To choose an Islander from the interface as a subject for a study the student needs to select their village and then select their house and then select the individual within the house. With variable numbers of houses and individuals in each of these layers the naive solution for choosing a ‘random’ sample will likely be biased. The design of the Island simulation is that it should run in real time. We continue the historical simulation, updating the population at the start of each month. In this way the underlying environment changes slowly over time with some Islanders dying and new ones being born. In December 2010 the Islanders even settled a new village that previously had not existed in the simulation.There is a fundamental tension in our design. We want the simulation to be realistic since we believe that will help students engage with the virtual environment. However there is a point at which realism becomes counterproductive towards our aims of engaging students in the role of statistical reasoning in scientific inquiry. For example, suppose we included a cause of death called Lung Cancer and made it so that Islanders with higher smoking levels were more likely to die from this disease. Students could collect data on smoking history and cause of death and look for this relationship but if they found an association it would probably not be surprising or interesting to them since it is the outcome they expect. They will not have discovered anything new by conducting their study. Instead of using real names for diseases we have thus tried to use poetic names wherever possible. These include Summer’s Pain , Diego’s Cough, Ruin and Jungle Sickness. One of these four is indeed modelled on lung cancer, including the association with smoking history, but now it is a more open question for students to explore. For example, what data do you need to collect to distinguish between these conditions and how can you convince somebody that you have identified ‘lung cancer’ on the Island? The virtual population described so far can be viewed as a framework for adding further simulations. The histories and images of the population give a broad context for experimental investigation. Returning to the original aim of this paper, we can thus enable students to obtain the various benefits of project work by adding the kinds of tasks which would provide the appropriate data to help learn statistical thinking. We have used this environment with an introductory statistics course for science students with around 1,200 students per year. They are asked to complete an experimental project to “demonstrate your understanding of the statistical methodology you have learned in the course”. One particular advantage of the online environment is that students can conduct their experimental work quite quickly towards the end of the semester. This means that they can have a clear statistical method in mind rather than the all too common practice of just collecting data without any plan of how to analyze it. Section 4 will show some examples of student work but we begin in this section with an overview of the design and mechanics of the experimental environment.A key feature of this innovation has been the involvement of students in its creation. From the outset we planned a two-phase curriculum design process for developing and using the Island in experimental projects. The first phase involved an assessment task where students had to prepare a research proposal with the Islanders as their subjects. For each student proposal that required an addition to the Island we began by searching for existing research on the topic. This gave plausible ranges for response variables as well as suggesting relationships that might be included in the simulation. As before there is tension between reality and fantasy here. Making the simulated processes perfectly match reality would be technically difficult and, as with the smoking and lung cancer example, may not actually be desirable. We felt it was important to keep students on their statistical toes by omitting some associations that they might expect to find while adding some other associations that would surprise them, though we did keep this at a low level. A better alternative for the long term is to add tasks that are somehow native to the Island. For example, in the first phase we added the fictional Dalpa Leaves and allowed Islanders to “chew lime-soaked dalpa leaves for ten minutes”. This was added as a control for chewing lime-soaked coca leaves but dalpa leaves were given their own effects that students could study independently in the future. The simulations used to generate the data that students observe involve a wide range of approaches. Our own earlier systems to simulate data for statistical exercises were based on standard models, such as using a linear model to generate outcomes based on parameters to which random Normal variability was added. In contrast the Island relies heavily on differential equations to capture changes in physiology at a more basic level and then links these with various statistical models as needed. While the population simulation moves in monthly steps, the experimental simulation updates at 30-second intervals. At this level each student has their own copy of the Island, tied to their login, so that changes they make to Islanders through experimental treatments are independent of changes made by other students. Similarly, such changes made by the students do not affect the underlying historical simulation. For example, a student can never kill an Islander through their actions since this would mean that they would then need a separate timeline in the historical simulation. Students conduct experiments by selecting an Islander and then allocating a task for them.