tectopulvinar

Communicating a Technical Topic

Last week I learned a very crucial lesson: know your audience. While I have given several neuroscience presentations during my college career, until last week I had never given one to an audience of students and faculty without background in the field. Speaking to a group of historians, statisticians, political scientists and writers, I learned firsthand the challenges involved in communicating highly technical neuroscience concepts to non-neuroscientists.

Going into this presentation, I was conscious of the complexity of my project. Most of the words I had to say were at least four syllables long. I had several slides with diagrams of overlapping neural pathways – some represented with blue lines, others with red, some with solid lines, others with dashed ones. It was a lot for me to explain, so I figured that it would be hard for someone unfamiliar with the concepts to follow. However, I underestimated just how challenging it would be for them to understand, and although I worked with my advisors to make the presentation more clear, it was still far too technical.

Looking back at this presentation and the feedback I received afterward, I have come to the conclusion that there were three main mistakes I made in communicating my thesis to this particular audience. My first mistake was failing to do a quick review of basic neuroscience principles at the beginning of the talk. Such a review would have contextualized my project and helped to catch my audience up to a level at which they could better understand my work. My rationale for not including this introduction was time; I wanted to maximize the time I spent talking about topics directly pertaining to my project. Therefore, I was hesitant to waste precious time on more foundational ideas. In reality, however, the time I saved by excluding this introduction was then wasted anyway by adding in other superfluous information elsewhere – my second mistake. I went on tangents thinking that extra information would help complete the picture of the issue I was addressing with my project. Because these tangents were also highly technical, however, all they did was confuse my audience further by distracting them from the more pertinent information. The final mistake I made was in the language I used. While I would argue that there were times when I could not get around using technical language to describe my project (e.g. the first time I introduced a key concept), I certainly did not need to rely on this technical language to the extent that I did. For example, rather than saying “temporal hemifield advantage” ten different times, I could have just said it once to define it and then subsequently referred to it as “a faster response to temporal stimuli.” Another, more simple alternative could have also been “the effect we are looking for.” In hindsight, using less technical language would have kept my audience more engaged with the result that they would have gotten more out of the talk overall.

If I had not made these three mistakes, my audience may have been able to follow my talk more easily. However, I do not regret making these mistakes. Although my audience may not have learned as much from this talk as I initially hoped, it was certainly a learning experience for me. Going forward, I know now that I will be more conscientious of the expertise of the people I am speaking to; I will choose to include specific information and language that will meet my audience’s need, with the hope of keeping them engaged throughout the whole lecture regardless of how technical the subject matter actually is.

Learn more about my project here.

Advertisements

Editing My Methods

A week ago, I thought I had successfully finished building the MATLAB script that will automatically present my dot stimuli. I had sorted out the last few bugs that were messing with the subject response key presses and had figured out how to fuse the subject’s vision so that each eye appeared to be receiving the same stimuli although, of course, they were not. I had tested it multiple times on myself without a hitch and even run the whole experiment successfully on a friend without any complications. Then, a graduate student in the lab suggested I make the dot stimuli slightly larger to account for the larger computer monitors I was using as part of my stereoscope. While this was a simple enough fix, it got me thinking: How do you know when you have finished editing your methodology?

There are a massive number of research papers out in the world today. Sure, that number shrinks as you specialize more and more – in my case, focusing on neuroscience papers pertaining to the superior colliculus and its role in visual cognition – but there are still a large number of papers, and thus methodologies, to choose from. Without a way for researchers to go back and comment on the validity and feasibility of their varying approaches to the same problem, it can be difficult to pick and choose what parts of their methodologies you should adopt in your own experiment.

For now, I have been mainly avoiding this problem by deferring to the opinions of those senior to me – my advisor, the graduate students and post-doctoral students in her lab. However, there may come a time when I am in a more senior position, myself, and have to advise others on the experiments that they are running. While I hope that by that time I have enough experience to advise them well, I also hope that by then there is a more objective way of distinguishing the relevancy of papers than just experience for even the most knowledgeable can make mistakes.

Seeking Programming Help

(This post was originally written on Friday, June 23.)

matlab

I have now been working on my thesis for five weeks. While in the first few weeks I was just reading papers in order to compile the beginnings of a literature review, recently I started tackling the MATLAB script that I will use to test my subjects. I have a feeling that this script will be one of the most challenging obstacles that I have to overcome in order to successfully complete my thesis.

I do not have a strong background in programming. Besides some brief experimentation with the language in high school, the only time I have used MATLAB to present stimuli was two years ago. At that time, I was working on the project of a post doc in the same lab that I am now doing my thesis. I only had to make a few minor changes to a script but that still took me a long time as I had to look up the appropriate way to write each command and then try to understand the bugs that invariably showed up. Over the ensuing two years, whatever moderate understanding of the language that I developed during that time seems to have vanished. Initially, this made me worried; there was no way for me to tackle this thesis without a functioning script. Out of this problem, however, emerged both a potential solution and an invaluable life lesson: It is okay to ask for help – I do not need to do this alone.

Since programming is such an integral part of conducting human-based behavioral and cognitive research, everyone in my lab has a lot of experience constructing a script; between all of the post docs and graduate students, they have written thousands of lines of code and worked through innumerable bugs. Consequently, I turned to them for help hoping that they could teach me how to program a script myself. As a skill that I may need to draw upon again and again in the future, I wanted to actually learn the language and how it can best be applied.

I am now in the midst of this learning process. Apart from getting more acquainted with MATLAB, this experience has also reacquainted me with the other members of Dr. Behrmann’s lab. Their support and technical expertise form a solid foundation upon which I have begun developing my own skills. Thus, while the prospect of constructing a script in MATLAB is still daunting, I now find myself looking forward to the challenge as I am no longer facing it alone.

Learn more about my project.

Temporal-nasal asymmetry: See it out of the corner of your eye

I have now been working on my thesis for two weeks. During that short amount of time, my project has already changed drastically. While I was warned by previous fellows that I need to be flexible because aspects of my thesis will change, I had not expected it to happen so early on. Then again, perhaps I am lucky that it did.

With my thesis, I am attempting to determine whether the superior colliculus plays a role in the subcortical visual processing of numerical stimuli. While subcortical visual processing has been well documented over the last few years through the presence of monocular advantage (i.e., an effect where people are faster at making same-different comparisons between sequentially presented stimuli when the stimuli are presented to one eye as opposed to both), the actual mechanism by which it works with respect to numerical stimuli is still not clear.

Although I only recently began the research for my thesis, I have already had to re-evaluate my methodology. Initially, I decided to target the superior colliculus using purple stimuli. Because of the short wavelength of purple light, purple images activate S-cones (short wavelength cones). Up until recently, S-cones were thought to not activate the superior colliculus. Thus, a failure to induce monocular advantage using purple stimuli would be indicative of the superior colliculus’ involvement in subcortical visual processing. As a technique used for decades to target the superior colliculus, this seemed like the best way to test its involvement. However, Hall & Colby (2014) found by recording the activation of individual neurons in the superior colliculus that the superior colliculus can be activated by S-cone-specific visual stimuli. This realization forced me to rethink my approach to the experiment.

A less commonly used technique to target the superior colliculus is temporal-nasal asymmetry. This technique takes advantage of the fact that the superior colliculus is part of the tectopulvinar pathway (a visual pathway that relies more on subcortical structures than the more cortically reliant geniculostriate visual pathway). In temporal-nasal asymmetry, stimuli are presented in either the temporal or nasal hemifields. Because of the asymmetry of the tectopulvinar pathway, temporal stimuli activate the superior colliculus more than nasal stimuli, in what is called a temporal hemifield advantage. Consequently, if stimuli that have previously uncovered monocular advantage result in a temporal hemifield advantage when presented in the temporal and nasal hemifields, this would suggest that the monocular advantage previously found may have been due to the superior colliculus.

My thesis has only just begun but my methods have already been altered. While such drastic changes so early on in the process are somewhat daunting, I look forward to pursuing this new direction. In the long run, I expect that it will be beneficial that I caught this problem now; it is easier to rectify a problem when it has only just begun to manifest rather than after everything has been completed.