Skip to content

Image analysis of phytoplankton and periphyton communities

While traveling home from the 41st annual Aquatic Toxicity Workshop in Ottawa, Ontario, Canada earlier this month, I began thinking about how much work I put into making oral and poster presentations at conferences and how sad it is that they are only shared with my scientific colleagues for one day during the conference. After that, these presentations just end up sitting on my laptop hard drive where no one can view them except for myself. Then I realized that many of these presentations could easily be posted online as blog posts. I’ve decided I’m going to start making blog posts from conference presentations that I’ve given over the last several years. The image below is a poster I presented at the annual meeting of the South-Central Regional Chapter of the Society of Environmental Toxicology and Chemistry in May 2013.Mudge et al. 2013 SETAC SC regional meeting Houston

Optimal alpha web calculator

Thanks to Leanne Baker and the R package ‘Shiny‘ from the makers of RStudio, the long-awaited, much anticipated optimal alpha web calculator is now up and running for t-tests, ANOVA, regression/correlation, and Chi-squared tests!

The optimal alpha web calculator can be accessed here:

Optimal alpha web calculator

Here’s a screenshot:

screenshot t-test

Try it out and let me know what you think!

Understanding, modeling and taking the focus off grades: My teaching philosophy

I’m working on a job application and I thought it would be a good idea to include a teaching philosophy. Here’s what I’ve come up with so far. I’d appreciate hearing any comments or suggestions you may have upon reading it.

Teaching philosophy

In the Information Age, students are no longer limited by the accessibility of information. My teaching objective is, therefore, to go beyond the presentation of information and encourage the development of real understanding in my students. More specifically, I intend to facilitate the understanding of (1) the relevance of information, (2) the relationships among different pieces of relevant information and (3) the applications of different sets of related information. Along with facilitating the development of a real understanding of course topics, I also believe it is important to teach students tools for evaluating their own understanding, because increasing understanding is far easier when knowledge gaps can be identified.

Science continually strives for a better understanding of nature, partly, if not exclusively, to increase our predictive capacity of the complex dynamics of the natural world. Predictive capacity is generated through the construction of models that seek to reduce the complexity of nature down to the components with the strongest influences. The construction of predictive models requires putting each of my teaching objectives into practice. Modeling requires the identification of the relative relevance of different pieces of information and an understanding of relationships among these pieces of information. The information is then applied for a particular predictive purpose and evaluation and validation of the predictive capacity of the model constitutes a tool for demonstrating our current level of understanding of a system. I intend to develop a strong understanding of course material in my students through the development, parameterization, and validation of predictive models that are applied to course concepts. I believe that many students studying environmental sciences do not receive enough training in modeling and are therefore uncomfortable with making and using models. I see this as a major problem. I believe that providing students with training and experience in constructing predictive models will enable students to apply a modeling perspective in their other courses and, eventually, in their future careers.

Students are often grade-focused, which can get in the way of understanding. Poor grades can discourage students from engaging themselves, as they find little reward for their effort. Good grades can also discourage students from trying harder than the minimum level that will result in a good grade. Because of this, I employ teaching methods designed to take student focus away from grades. (1) I prefer to give more descriptive feedback and fewer grades. This feedback consists of written comments and individual feedback sessions, in addition to self-evaluation and peer-evaluation techniques. When giving feedback, I believe it is equally important to provide students with recognition of things done well in addition to providing suggestions for improvement. (2) Whenever possible, I assign projects that can have real-world impacts that students can become engaged in.

Virtually anything that students could wish to learn about is available over the internet, if not for free then usually for a fee much smaller than the cost of a university course. I hope that my teaching provides value added learning experiences for students, such that it is worth students’ time and money to learn through taking my course than by independent online learning. Accomplishing this requires that I (and the content I present) be engaging and accessible and that I take full advantage of the capabilities for in-class activities and discussions that are currently more difficult to achieve online.

 

Has Occam’s razor shredded ecology to bits?

At its most basic level, ecology attempts to explain the richness and abundances of species over space and time.

I feel that ecology has been notoriously poor at achieving this goal in most situations. It is not uncommon for fisheries to collapse, for pest control measures to be unsuccessful and for species conservation efforts or re-introductions to fail. Although there are often political and/or economic factors contributing to these failures of applied ecology, poor and/or contradictory ecological predictions frequently also play a role.

Compared to other scientific disciplines, ecology has disappointingly few predictive theories that are both practically useful and broadly applicable. Instead, ecology is highly fragmented and our explanations for the richness and abundances of species over space and time tend to be both taxonomically and geographically specific. The common excuse for this situation is that nature is complex. Few dispute this. However, why should the complexity of nature be an adequate excuse for favouring simple, system-specific explanations of species richness and/or abundance patterns? In practice, this leads to uncertainty when attempting to apply ecological principles to new, previously uninvestigated systems.

Why aren’t we looking for a grand, universal model for predicting species richness or species abundances over space or time? It can’t be possibly be for lack of data. I do suspect that any initial attempts at a universal model for predicting species richness or abundance over space or time would be complex and would also likely have both low predictive capacity and low universality, but I would argue that this is also true for our current system-specific models.

In taking a reductionist view of ecology, we have parsed it out into a multitude of sub-disciplines and found that different things are important predictors of species richness or abundance in different sub-disciplines, and we’ve stopped there. It seems that we’ve lost sight of “understanding the whole” and instead are sprawling out into examining more and more parts from different perspectives. I believe that favouring small-scale simplicity in this way is strongly impeding the unified, integrated progression of the discipline as a whole.

Developing universal models for predicting species richness or abundances over space or time (expressed in terms of information we have the ability to measure) would be a very daunting task. I do, however, think it is sorely needed in order for the highly fragmented field of ecology to provide us with a more complete understanding of our natural world.

“It can scarcely be denied that the supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.” – Albert Einstein, in “On the Method of Theoretical Physics,” the Herbert Spencer Lecture, Oxford, June 10, 1933.

Or, more simply:

“Everything should be as simple as it can be but not simpler.” – Robert Sessions, in “How a ‘Difficult’ Composer Gets That Way,” The New York Times,  January 8, 1950.

Please feel free to chime in with any ideas of potential starting points for universal ecological models, or with any measurable values that might make important parameters in such models.

Frequentists vs. Bayesians (and a comment about the optimal alpha approach)

Frequentists vs. Bayesians

If the frequentist had used the optimal alpha approach and calculated an optimal alpha for a scenario of unequal prior probabilities of null and alternate hypotheses, he/she would have agreed with the Bayesian.

The optimal alpha approach using unequal relative costs of error would have also resulted in the frequentist agreeing with the Bayesian. The cost of error Type I error (falsely concluding the sun has gone nova) would be $50 because he/she would have lost the bet.  The cost of Type II error (falsely concluding the sun has not gone nova) is negligible because there’s really no advantage to knowing the sun has just exploded since we’d soon all be dead. If Type I errors are more serious than Type II errors, the resulting optimal alpha that minimizes costs of errors would be much smaller than 0.05, leading the frequentist to the conclusion that the sun hasn’t gone nova.

Are ecological modeling papers just flashes in the pan?

A few months ago, I came across an interesting tool for comparing the citations of different categories of papers within a research field (within a country), while looking up some journal impact factors on the SCImago Journal & Country Rank website. When I used this tool for 2009-2010 United States papers in the field of environmental science, here’s what I got:

Apparently ecological modeling papers are more likely to be cited by other papers, and they get more citations on average, compared to all other subject categories in environmental science. The size of the circles corresponds to the number of papers published in that subject category, so the other thing that stood out to me was how few ecological modeling papers are published compared to some of the other categories. Basically, if ecological modeling papers are so well cited, why aren’t more people doing this kind of research? I decided to dig a bit deeper into this.  If we go back one year earlier, here’s the graph of United States environmental science paper categories from 2008-2009:

Ecological modeling papers published in 2008-2009 don’t have the highest average number of citations anymore. They’re beat by papers about global and planetary change. Let’s go back one more year and look at United States environmental science paper categories from 2007-2008:

These older ecological modeling papers published in 2007-2008 are starting to have citation records closer to the other categories. They get beat again by papers about global and planetary change, this time in both average number of citations and the % of papers with at least 1 citation. If we go back one more year to US environmental science paper categories from 2006-2007, here’s what we get:

For papers published in 2006-2007, papers about global and planetary change and also general ecology papers now have higher averages of citations per document than ecological modeling papers. There seems to be a pattern emerging here. So if I go back 10 years, to US environmental science paper categories from 2001-2002, here’s the result:

What happened to ecological modeling papers? Well for one thing, their circle colour changed from tan to purple for some unknown reason. It also looks like people have basically almost stopped citing the ecological modeling papers published from 2001-2002, while many other paper types continue to get steady citations and end up coming out ahead in the long run.

Take home message: If you do environmental science research and you want to write a paper that gives you the best chance of getting a few quick citations, write an ecological modeling paper. However, if you’re looking to write a paper that will remain a valued piece of literature in the long-run, an ecological modeling paper is less likely to accomplish this.

There are probably several factors that lead to the slowing down of the citation rate of ecological modeling papers over time, relative to other environmental science paper categories. Please leave a comment if you’ve got any ideas on what some of the most important ones might be.

Are academic conferences obsolete in the information age?

Of the 4 academic conferences I’ve attended in the past 2 years, each one of them has resulted in a publication related to the content I presented at the conference. So, clearly, I’ve gotten some benefit from giving presentations at academic conferences.

What I’ve been wondering lately is whether I could have gotten the same benefits cheaper and easier without ever leaving my office.

I can think of 4 different kinds of benefits that researchers gain from attending academic conferences and each benefit seems like it’s becoming increasingly easy to gain online.

Benefit 1: Keeping up to speed on colleagues’ latest research

This can easily be accomplished by regular visits to journals’ online early-view web pages, and by subscribing to the blogs and/or email lists of journals, research societies and research colleagues.

Benefit 2: Communicating your latest research to your colleagues

Over the past year I have presented at two conferences and published two blog posts related to an article I recently published in PLOS ONE. PLOS ONE provides metrics for each article that allow anyone to track the page views and downloads of an article over time. After each conference, I could barely detect any jump in page views or downloads in my PLOS ONE article. However, after each blog post, I saw a definite boost in both page views and article downloads that lasted for weeks after making each blog post.

This suggests:

Online communication of new research is more effective than conference presentations on new research findings and/or I’m not very good at making conference presentations… I hope the former is the case.

Benefit 3: Networking

Networking can mean many things.

What I call ‘professional networking’ involves creating new professional connections by showing you’d be an intelligent person to collaborate with. My feeling is that this type of networking can be easily accomplished online by making regular intelligent comments on research blogs and listservs.

What I’m calling ‘social networking’ (not to be confused with facebooking) involves creating new connections among colleagues by showing you’d be an easy/fun person to collaborate with. This type of networking is admittedly easier by attending social events at conferences.

My personal preference is to gain network connections by showing that I’d be a valuable person to collaborate with, rather than gaining connections by showing I’d be an easy person to collaborate with (although I do also like to think I’m an easy person to collaborate with).

Benefit 4: Research-focused events

Conferences still have the edge over the internet with respect to offering professional workshops and meetings on important research topics. However, these aspects of conferences are typically not the focus. They are often scheduled outside the main dates of the conference, they sometimes cost extra to attend and as a result, they are usually poorly attended compared to the conference presentation sessions. Also, the capacity for hosting meetings and workshops online is increasing. In a few years it may be just as easy to facilitate these events online as at an academic conference.

Wrapping things up

Conferences have traditionally been an important forum for ‘confer’ring with research peers, but the internet may be making them less necessary than they once were. Most of us like to travel and socialize with like-minded people, but can we really justify the time and expense of academic conferences for these reasons alone?

Please let me know what you think.

If you agree that conferences are becoming obsolete, feel free to provide any tips on how you gain conference-type benefits without attending conferences.

If you feel that academic conferences are still essential in the internet age, please comment on why you have this opinion, in the hopes of restoring my enthusiasm for attending academic conferences in the future.

Notes on Learnings About Data Analysis

Smile! You’re at the best WordPress.com site ever

Early Career Ecologists

Questioning, Researching, and Communicating our Science

Ecologically Orientated

...with frequent whatnot

The Lab and Field

Science, people, adventure

The Contemplative Mammoth

ice age ecology, early career academia, and diversity in STEM

Just Simple Enough: The Art of Mathematical Modelling

Exploring the secrets of successful mathematical models

Jabberwocky Ecology

Ethan White and Morgan Ernest's blog for discussing issues and ideas related to ecology and academia.

theoretical ecology

notes from ecology, biogeography and evolution by Florian Hartig

Dynamic Ecology

Multa novit vulpes

regression to the mean

my opinion on all things.

Sociobiology

So you want to be a biology professor

Biodiversity Revolution

Thoughts from the vanguard of biodiversity research

Arthropod Ecology

Writings about arthropod ecology, arachnids & academia at McGill University

Statisfaction

I can't get no