I’ve been reading a couple of old books of book reviews by Anthony Burgess. Lots of great stuff. He’s a sort of Chesterton with a conscience, for example in this appreciation of Uncle Tom’s Cabin: As for Tom’s forgiving Christianity—‘O, Mas’r! don’t bring this great sin on your soul! It will hurt you more than […]

I originally gave this post the title, “Stigler: The Changing History of Robustness,” but then I was afraid nobody would read it. In the current environment of Move Fast and Break Things, not so many people care about robustness. Also, the widespread use of robustness checks to paper over brittle conclusions has given robustness a […]

tl;dr: Their 95% interval for the infection rate, given the data available, is [0.7%, 1.8%]. My Bayesian interval is [0.3%, 2.4%]. Most of what makes my interval wider is the possibility that the specificity and sensitivity of the tests can vary across labs. To get a narrower interval, you’d need additional assumptions regarding the specificity […]

Kevin Lewis pointed me to this article. It seemed kinda familiar, I took a look at the abstract, and I realized . . . I reviewed this article for the journal! Here was my referee report: The paper seems fine to me. I have only two minor comments, both relating to the abstract. 1. I […]

We’ve had lots of discussions here and elsewhere online about fundamental flaws in statistics culture: the whole p-value thing, statistics used for confirmation rather than falsification, corruption of the pizzagate variety, soft corruption in which statistics is used in the service of country-club-style backslapping, junk science routinely getting the imprimatur of the National Academy of […]

1. The news! Zad’s cat, pictured above, is not impressed by this bit of cargo-cult science that two people sent to me: No vaccine or effective treatment has yet been found for people suffering from COVID-19. Under the circumstances, a physician in Kansas City wonders whether prayer might make a difference, and he has launched […]

Anthony Fowler wrote a wonderful op-ed. You have to read the whole thing, but let me start with his most important point, about “the temptation to overclaim” in social science: One study estimated the economic value of the people spared through social-distancing efforts. Essentially, the authors took estimates from epidemiologists about the number of lives […]

The other day we posted some Stan models of coronavirus infection rate from the Stanford study in Santa Clara county. The Bayesian setup worked well because it allowed us to directly incorporate uncertainty in the specificity, sensitivity, and underlying infection rate. Mitzi Morris put all this in a Google Collab notebook so you can run […]

See here. Please share your reactions and suggestions in comments. I’ll be talking with Seth Flaxman tomorrow, and we’d appreciate all your criticisms and suggestions. All this is important not just for Italy but for making sensible models to inform policy all over the world, including here.

Carlos Ungil writes: The IHME released an update to their model yesterday. Using now a better model and taking into account the relaxation of mitigation measures their forecast for US deaths has almost doubled to 134k (95% uncertainty range 95k-243k). My [Ungil’s] charts of the evolution of forecasts across time can be found here. I […]

What’s new? The new and shiny reduce_sum facility released with Stan 2.23 is far more user-friendly and makes it easier to scale Stan programs with more CPU cores than it was before. While Stan is awesome for writing models, as the size of the data or complexity of the model increases it can become impractical […]

The University of Washington in Seattle is a big place. It includes the Institute for Health Metrics and Evaluation (IHME), which has produced a widely-circulated and widely-criticized coronavirus model. As we’ve discussed, the IHME model is essentially a curve-fitting exercise that makes projections using the second derivative of the time trend on the log scale. […]

Part 1 Andrew Lilley, Gianluca Rinaldi, and Matthew Lilley write: You might be familiar with a recent paper by Correira, Luck, and Verner who argued that cities that enacted non-pharmaceutical interventions earlier / for longer during the Spanish Flu of 1918 had higher subsequent economic growth. The paper has had extensive media coverage – e.g. […]

Robin Lee writes: To check for linearity and homoscedasticity, we are taught to plot residuals against y fitted value in many statistics classes. However, plotting residuals against y fitted value has always been a confusing practice that I know that I should use but can’t quite explain why. It is not until this week I […]

By default, ggplot expands the space above and below the x-axis (and to the left and right of the y-axis). Andrew has made it pretty clear that he thinks the x axis should be drawn at y = 0. To remove the extra space around the axes when you have continuous (not discrete or log […]

David Rohde points us to this new seminar series that has the following description: Machine learning is changing the world we live in at a break neck pace. From image recognition and generation, to the deployment of recommender systems, it seems to be breaking new ground constantly and influencing almost every aspect of our lives. […]

John Fund writes: [Imperial College epidemiologist Neil] Ferguson was behind the disputed research that sparked the mass culling of eleven million sheep and cattle during the 2001 outbreak of foot-and-mouth disease. He also predicted that up to 150,000 people could die. There were fewer than 200 deaths. . . . In 2002, Ferguson predicted that […]

After getting 80 zillion comments on that last post with all that political content, I wanted to share something that’s purely technical. It’s something Bob Carpenter wrote in a conversation regarding implementing algorithms in Stan: One thing we are doing is having the matrix library return more expression templates rather than copying on return as […]

Ivan Oransky points us to this amusing retraction of a meta-analysis. The problem: “Standard errors were used instead of standard deviations when using data from one of the studies”! Actually, I saw something similar happen in a consulting case once. The other side had a report with estimates and standard errors . . . the […]

The engineering librarians at the Art, Architecture & Engineering Library are pleased to announce that some engineering textbooks for the Fall 2019 semester are available to you in full text online through Library subscriptions.

Please join the engineering librarians for a series of 30-minute workshops at the Duderstadt Center this semester, on topics such as Data Visualization, document creation with LaTeX, and finding funding for your research.

Starting November 1 the University of Michigan Library has access to the new version of chemistry database SciFinder — SciFinder-n. In fact, both will be accessible to our users with the same login and password information.

The engineering librarians at the Art, Architecture & Engineering Library are pleased to announce that some engineering textbooks for the Winter 2020 semester are available to you in full text online through Library subscriptions.

Join us on Monday, Feb. 24th at noon in the Materials Collection (AAEL 2nd Floor) to hear about the research of Prof. Brian Love (MSE). He will speak about his research on extracting fibrous plant residues from agricultural biomasses, such as coffee, to make other continuous fibers.

Following Governor Whitmer's executive order to close places of public accommodation, the Duderstadt Center and all U-M Library buildings have closed to all visitors as of Monday, March 16 at 3pm.

In forest reserves of the U.S. Pacific Northwest, management objectives include protecting late-seral habitat structure by reducing the threat of large-scale disturbances like wildfire. We simulated how altering within and among-stand structure with silvicultural treatments of differing intensity affected late-seral forest (LSF) structure and fire threat (FT) reduction over 30 years in a 6070-ha reserve. We then evaluated how different financial requirements influenced the treatment mix selected for each decade, the associated effects on FT reduction and LSF structure in the reserve, and treatment costs. Requirements for treatments to earn money (NPV+), break even (NPVO), or to not meet any financial goal at the scale of the entire reserve (landscape) affected the predicted reduction of FT and the total area of LSF structure in different ways. With or without a requirement to break even, treatments accomplished about the same landscape level of FT reduction and LSF structure. Although treatment effects were similar, their associated net revenues ranged from negative $1 million to positive $3000 over 30 years. In contrast, a requirement for landscape treatments to earn money ($0.5 to $1.5 million NPV) over the same period had a negative effect on FT reduction and carried a cost in terms of both FT reduction and LSF structure. Results suggest that the spatial scale at which silvicultural treatments were evaluated was influential because the lowest cost to the reserve objectives was accomplished by a mix of treatments that earned or lost money at the stand level but that collectivel broke even at the landscape scale. Results also indicate that the timeframe over which treatments were evaluated was important because if breaking even was required within each decade instead of cumulatively over all three, the cost in terms of FT reduction and LSF structure was similar to requiring landscape treatments to earn $0.5 million NPV.

This publication provides estimates of total softwood and hardwood harvests by region and owner for the United States from 1950 to 2002. These data are generally not available in a consistent fashion and have to be estimated from state-level data, forest resource inventory statistics, and production of forest products. This publication describes the estimation process and documents the various assumptions. These estimates have been used for the past three decades in the periodic USDA Forest Service timber assessments.

Millions of miles of highway crisscross the United States. Highways fragment the landscape, affecting the distribution of animal populations and limiting the ability of individuals to disperse between those populations. Moreover, animal-vehicle collisions are a serious hazard to wildlife, not to mention people.

A century of effective fire suppression has radically transformed many forested landscapes on the east side of the Cascades. Managers of dry forests critically need information to help plan for and implement forest restoration. Management priorities include the stabilization of fire regimes and the maintenance of habitat for the northern spotted owl and other old-forest associates.

In the mid-1990s, forest managers on the Malheur National Forest were concerned about their prescribed fire program. Although they have only a few weeks of acceptable conditions available in the spring and fall, they were worried that spring-season prescribed burning might be exacerbating black stain root disease and having negative effects on understory plants.

Definitive results from the Starkey Project's first decade (1989-99) have given managers defensible options for managing roads, timber production, and range allotments in relation to elk, deer, and cattle. Study results have prompted changes in policies, management standards and guidelines, hunting regulations, and timber sale planning throughout western North America.

New equipment, technology, and marketing efforts have allowed Alaska's wood products producers to consider opportunities previously unavailable to them. Until recently, the primary product produced by Alaska firms was rough, unseasoned lumber sold primarily within local markets. Given the purchase and installation of new drying and planing equipment, Alaska producers can now enter domestic and export markets for a variety of secondary wood products. Previously underutilized species, such as red alder (Alnus rubra Bong.), paper birch (Betula papyrifera Marsh.), and Alaska yellow-cedar (Chamaecyparis nootkatensis (D. Don) Spach) are also gaining in popularity and market potential. A detailed knowledge of species preferences for Alaska lumber, across business types and geographic regions, will be essential if Alaska producers are to be competitive.

The Pacific Northwest Research Station launched a Biodiversity Initiative to assist natural resource professionals in integrating complex biodiversity concepts into natural resource management processes. We canvassed clients from various affiliations to determine the main challenges they face in biodiversity management, to define their information needs, and to understand how best to deliver biodiversity information within a collaborative framework. The biodiversity management challenges that emerged included (1) the lack of well-defined biodiversity management policies, (2) understanding and quantifying the interaction effects between a number of factors (e.g., disturbance types, management practices) and biodiversity, (3) the lack of applied biodiversity monitoring strategies, (4) difficulty in locating and accessing biodiversity information, and (5) balancing conflicting values relating to biodiversity. We also list the biodiversity information product needs of clients, as well as preferred technology transfer methods, and we discuss the future direction of the Biodiversity Initiative.

A century of fire suppression has resulted in dense fuel loads within the dry pine forests of eastern Oregon . To alleviate the risk of stand-replacing wildfire, forest managers are using prescribed fire and thinning treatments. Until recently, the impact of these fuel treatments on soil productivity has been largely unknown. Such information is essential for making sound management decisions about the successful reintroduction of fire to the ecosystem to retain biodiversity of soil fungi and achieve the desired future condition of large ponderosa pines with low fuel loads. In a recent pair of studies, led by researchers at the PNW Forestry Sciences Laboratory in Corvallis, Oregon, novel molecular techniques were utilized to investigate the response of soil ecosystems to prescribed burning and thinning. The research compared impacts of the season of burn and various combinations of fuel-reducing treatments. Results suggest that overly severe fires can damage soil productivity and that less intense fires can be used to gradually reduce accumulations of fuel. The findings are currently being implemented in decisions about forest management and contribute important new information to the science.

My Fuel Treatment Planner (MyFTP) is a tool for calculating and displaying the financial costs and potential revenues associated with forest fuel reduction treatments. It was designed for fuel treatment planners including those with little or no background in economics, forest management, or timber sales. This guide provides the information needed to acquire, load, and begin to use MyFTP.

The socioeconomic monitoring report addresses two evaluation questions posed in the Northwest Forest Plan (the Plan) Record of Decision and assesses progress in meeting five Plan socioeconomic goals. Volume I of the report contains key findings. Volume II addresses the question, Are predictable levels of timber and nontimber resources available and being produced? It also evaluates progress in meeting the goal of producing a predictable level of timber sales, special forest products, livestock grazing, minerals, and recreation opportunities. The focus of volume III is the evaluation question, Are local communities and economies experiencing positive or negative changes that may be associated with federal forest management? Two Plan goals are also assessed in volume III: (1) to maintain the stability of local and regional economies on a predictable, long-term basis and, (2) to assist with long-term economic development and diversification to minimize adverse impacts associated with the loss of timber jobs. Progress in meeting another Plan goal-to promote agency-citizen collaboration in forest management-is evaluated in volume IV. Volume V reports on trends in public values regarding forest management in the Pacific Northwest over the past decade, community views of how well the forest values and environmental qualities associated with late-successional, old-growth, and aquatic ecosystems have been protected under the Plan (a fifth Plan goal), and issues and concerns relating to forest management under the Plan expressed by community members. Volume VI provides a history of the Northwest Forest Plan socioeconomic monitoring program and a discussion of potential directions for the program.

Decayed wood plays many critical roles in forest ecosystems. Standing dead trees, called snags, provide habitat for a suite of wildlife, including several species of birds, insects, bats, and other mammals. Down wood provides wildlife habitat and performs ecosystem services such as releasing humus, nitrogen, and phosphorus into the forest soil, storing pockets of moisture, and stabilizing soil on slopes. Root wads, tree stumps, hollow trees, and partially dead trees also perform important ecological roles as wildlife habitats and sources of soil organic matter. DecAID Advisor is an on-line decision-aiding system to help managers plan for wood decay elements for biodiversity in forests of Washington and Oregon. DecAID Advisor is a statistical "meta-analysis" and synthesis of a vast amount of wildlife and inventory data. It does not make decisions for managers, but instead, DecAID Advisor advises on size and amount of snags, down wood, and other wood decay elements to meet management objectives and to help set those objectives by forest type and structural condition class. It is the first decision-aiding tool of its kind, given its scope of species, inventory data, and topics provided.

Nontimber forest products from southern Alaska (also called special forest products) have been used for millennia as resources vital to the livelihoods and culture of Alaska Natives and, more recently, as subsistence resources for the welfare of all citizens. Many of these products are now being sold, and Alaskans seek additional income opportunities through sustainable harvest and manufacture of such forest resources. We discuss the unique legal, regulatory, land tenure, geographic, vegetation, and climatic context that southern Alaska presents for marketing nontimber forest products; summarize the various species and types of products being harvested; and consider the marketing challenges and opportunities new entrepreneurs will encounter. The information and resources we provide are intended to enhance income opportunities for all Alaskans, while sustaining the organisms harvested, respecting traditional activities, and ensuring equitable access to resources.

The Pacific Northwest is one of the world's major timber-producing regions, and its capacity to produce wood on a sustained-yield basis is widely recognized. Nonetheless, there has been increasing public interest in assuring that forests are being sustainably managed, as well as a desire by landowners to demonstrate their commitment to responsible stewardship.

The Community Socioeconomic Information System (CSIS) is a tool that allows users to retrieve 1990 and 2000 U.S. census data to examine conditions and trends for communities in western Washington, western Oregon, and northern California. The tool includes socioeconomic data for 1,314 communities in the entire region, including incorporated and unincorporated places. The tool delivers socioeconomic data using mapping and database features. In addition to providing data for one community, the tool produces community-level data at a variety of scales, including communities in areas surrounding Forest Service and Bureau of Land Management lands, all communities in the Northwest Forest Plan (NWFP) region, and communities within planning provinces within the NWFP region. One feature allows users to customize community data by creating boundaries and socioeconomic data for group of selected communities. The CSIS tool was designated to increase the usefulness of socioeconomic information at the small scale. Typically community socioeconomic assessments use U.S. census designations called census places. However, census places only represent a portion of the rural population. The CSIS uses a smaller unit of analysis (block groups) that we have aggregated to represent contiguous communities across the landscape, thereby representing the entire population. Community data can be printed as reports with graphs and tables, queried within an Access database, mapped and queried as geographic information system (GIS) data within ArcExplorer (a free GIS software included), exported as a table for use in Excel, or exported as GIS data for use in ArcGIS. The tool has features that allow users to locate communities by county or state and become familiar with local geography. The CSIS includes GIS data, such as major land ownerships, political boundaries, and physical landscape features. Applications produce maps that can be printed for specific communities showing community boundaries, water features, roads, metropolitan areas, community population centers, public land ownership, census places, planning provinces, counties, and state boundaries. Or, using the spatial data provided on the CD and ArcExplorer, users can produce custom maps.

The Northwest Forest Plan (the Plan) is a large-scale ecosystem management plan for federal land in the Pacific Northwest. Marbled murrelet (Brachyramphus marmoratus) populations and habitat were monitored to evaluate effectiveness of the Plan. The chapters in this volume summarize information on marbled murrelet ecology and present the monitoring results for marbled murrelets over the first 10 years of the Plan, 1994 to 2003.

WestProPlus is an add-in program developed to work with Microsoft Excel to simulate the growth and management of all-aged Douglas-fir-western hemlock (Pseudotsuga menziesii (Mirb.) Franco-Tsuga heterophylla (Raf.) Sarg.) stands in Oregon and Washington. Its built-in growth model was calibrated from 2,706 permanent plots in the Douglas-fir-western hemlock forest type in Oregon and Washington. Stands are described by the number of trees per acre in each of nineteen 2-in diameter classes in four species groups: Douglas-fir, other shadeintolerant species, western hemlock, and other shade-tolerant species. WestProPlus allows managers to predict stand development by year and for many decades from a specific initial state. The simulations can be stochastic or deterministic. The stochastic simulations are based on bootstrapping of the observed errors in models of stand growth, timber prices, and interest rate. When used in stochastic simulations, this bootstrap technique simulates random variables by sampling randomly (with replacement) from actual observations of the variable, rather than from an assumed distribution. Users can choose cutting regimes by specifying the interval between harvests (cutting cycle) and a target distribution of trees remaining after harvest. A target distribution can be a reverse-J-shaped distribution or any other desired distribution. Diameterlimit cuts can also be simulated. Tabulated and graphic results show diameter distributions, basal area, volumes by log grade, income, net present value, and indices of stand diversity by species and size. This manual documents the program installation and activation, provides suggestions for working with Excel, and gives background information on West-ProPlus's models. It offers a comprehensive tutorial in the form of two practical examples that explain how to start the program, enter simulation data, execute a simulation, compare simulations, and plot summary statistics.

This problem analysis describes a variety of human-resource interaction issues and identifies related social science research and development needs that serve as the foundation for the Alaska Communities and Forest Environments Team within the Pacific Northwest Research Station. The document lays out a research agenda that focuses on understanding relations between human communities and natural resources.

Some federal forest managers working in late-successional reserves find themselves in a potential no-win situation. The Northwest Forest Plan requires that the reserves be protected from large-scale natural and human disturbances while simultaneously maintaining older forest habitat.

The Community Socioeconomic Information System (CSIS) is a tool that allows users to retrieve 1990 and 2000 U.S. census data to examine conditions and trends for communities in western Washington, western Oregon, and northern California. The tool includes socioeconomic data for 1,314 communities in the entire region, including incorporated and unincorporated places.

Currently, all treated lumber used in Alaska is imported from the 48 contiguous states and Canada because there are no wood-treating facilities in Alaska. This report explores conventional and alternative wood-treating methods and reviews previous studies and laboratory tests on treated wood. In investigating wood treatment as a possible processing option for Alaska forest products manufacturers, the double-diffusion method of using sodium fluoride followed by a copper sulfate appeared to be the most advantageous approach. This method of treating wood was identified because it can be used to treat freshly cut or green wood. This was an important factor to consider, owing to the limited drying capacity in Alaska. Little information was available as to the chemical retention after treating and its resistance to leaching.

Integrated research is about achieving holistic understanding of complex biophysical and social issues and problems. It is driven by the need to improve understanding about such systems and to improve resource management by using the results of integrated research processes. Traditional research tends to fragment complex problems, focusing more on the pieces of problems rather than the whole that comprises multiple interrelationships and interactions. The outcome is that a lot is known about the parts (e.g., recreation, fish, and wildlife) but relatively little about how they are interrelated. There seems to be general agreement that integrated questions must drive the search for integrated understanding, but tradition, inertia, institutional culture,budgets, training, and lack of effective leadership foster reductionism (at worst) or minimal degrees of integration (at best) rather than any substantial, sustainable effort toward integrated research. In this paper, a phased approach to framing integrated research questions and addressing the substantial barriers that impede integrated efforts are discussed. A key conclusion is that to make any significant progress toward comprehensive integrated research will require more than rhetoric. Progress must begin with more effective leadership throughout various levels of research organizations.

This report examines socioeconomic changes that occurred between 1990 and 2000 associated with implementation of the Northwest Forest Plan (the Plan) in the Olympic National Forest in western Washington. We used a combination of quantitative data from the U.S. census and the USDA Forest Service, historical documents, and interviews from Forest Service employees and members of three case study communities-Quilcene, the Lake Quinault area, and the Quinault Indian Nation. We explore how the Plan affected the flow of socioeconomic benefits associated with the Olympic National Forest, such as the production of forest commodities and forest-based recreation, agency jobs, procurement contract work for ecosystem management activities, grants for community economic assistance, payments to county governments, and opportunities for collaborative forest management. The greatest change in socioeconomic benefits derived from the forest was the curtailment of timber harvest activities. This not only affected timber industry jobs in local communities, but also resulted in declining agency budgets and staff reductions. Mitigation efforts varied. Ecosystem management contracts declined and shifted from labor-intensive to equipment-intensive activities, with about half of all contractors from the Olympic Peninsula. Economic assistance grants benefited communities that had the staff and resources to develop projects and apply for monies, but provided little benefit to communities without those resources. Payments to counties served as an important source of revenue for rural schools and roads. We also examine socioeconomic changes that occurred in the case study communities, and the influence of forest management policy on these changes. Between 1990 and 2000 all three communities showed a decrease in population, an increase in median age, a decline in timber industry-related employment, and an increase in service-industry and government jobs. Quilcene's proximity to the larger urban centers has attracted professional and service industry workers that commute to larger economic hubs. Lake Quinault area residents are increasingly turning to tourism, and its growing Latino population works in the cedar shake and floral greens industries. For the Quinault Indian Nation, employment in tribal government and its casino has helped offset job losses in the fishing and timber industries. Many changes observed in the communities were a result of the prior restructuring of the forest products industry, national economic trends, and demographic shifts. However, for Quilcene and Lake Quinault, which were highly dependent on the national forest for timber and served as Forest Service district headquarters, the loss of timber industry and Forest Service jobs associated with the Plan led to substantial job losses and crises in the economic and social capital of these communities.

This case study examines the socioeconomic changes that took place between 1990 and 2000 in and around lands managed by the Bureau of Land Management (BLM) Coos Bay District in southwestern Oregon for purposes of assessing the effects of the Northwest Forest Plan (the Plan) on rural economies and communities in the Coos Bay region.

Projections of Alaska timber products output, the derived demand for logs and chips, and timber harvest by owner are developed by using a trend-based analysis.