Bad at statistics? Don’t test me.

Recently I was in a meeting with one of my supervisors, already feeling out of my depth as we were editing my first paper. I nodded obediently as it was suggested that I should use a non-parametric test instead of what I had done.  Then I was asked a question that startled me out of my comfortable yes man role,

since the environmental variables are not normally distributed which test should you use for the discontinuous data?“.

ungh…err..” I valiantly replied

you have done statistics courses, right?

I conceded I had but I wasn’t sure what test to use*

I had undertaken statistics courses, several in fact, but I have never been particularly good at retaining abstract ideas unless I could practically applying them.

As I sweated through the meeting I berated myself, If I know I’m no good at statistics and maths, why oh why didn’t I take more courses to improve before undertaking a PhD?!

Well the answer was obvious, I wanted to pass.

If a student wants to achieve high scores and have a chance at a scholarship or to get into postgraduate study, there is a strong incentive to only choose subjects they are good at. Choosing subjects a student finds difficult has limited short term academic benefit aside from personal development, as there is no difficulty bonus points for taking harder subjects. Choosing a harder subject will also require spending more time on the perceived harder subject to pass, which will consequently siphon time from other subjects impacting those scores.

So how can these costs be overcome?

Potentially a different system of subject selection is required. Although grading subjects has it’s uses and provides incentive to work hard, it is also a part of the problem. To overcome this, universities could consider introducing a system whereby students can nominate at least one subject that is not graded, but purely marked on attendance.

This would provide an incentive for students to choose subjects that they may be interested in, or know they may eventually need, but are hesitant to choose these subjects as they may require too much time commitment, or the student fears they have little chance of passing. This system could also facilitate cross-disciplinary learning  as it would allow the selection of a subject purely for  interest, rather than concern for learning the key areas of a students course.

Now if you excuse me, I need to read this statistics book.

*(It was a Mann Whitney U test for those playing along at home)

Posted in Uncategorized | Tagged , , | Leave a comment

Give ‘em flowers

The backyard vegetable garden is a lot like a farm in  that they are both optimised for  production. When starting a veggie garden the first action that people take is to clear the area of any grass or plants that could compete with the planted vegetables. This cleared area is also often away from the  rest of the garden so its not in the shade and gets plenty of sunlight.

While this garden design minimises the competition faced by vegetable plants for nutrients, water, light and thereby maximising their growth, it does have it’s drawbacks. Removing and isolating the veggie garden from a diversity of plants (such as your lavender, bottle brush and roses) can discourage beneficial insects from visiting.

Many insects in the hymenopteran family (wasps) feed either on garden insect pests or destroy the eggs of pest species through parasitism. This Native beecan be a great benefit in  controlling the amount of common garden pests such as caterpillars or aphids. However, most of these wasps at some point in their life cycle require nectar and pollen, the very resources that were removed when you cleared those other plants for your garden! In this respect a veggie garden would be better off being within the rest of the garden plants, least surrounded by flowers or even inter-planted with ornamental plants.

Studies on farms and plantations have had success in increasing beneficial wasp numbers by adding flowering plants between crop rows. Typically the flower alyssum was used in these studies (which can make a nice hedge around gardens) due to its prolific flowering, however any flower may have this affect.

So having flowers near to a garden year round can keep these beneficial wasps in higher numbers year round. Additionally, a constant supply of flowers can also be attractive to pollinators such as bees and many flies, while leaf litter dropped by ornamental plants can be attractive to decomposing insects like worms which breakdown the leaf litter and help aerate the soil.

These factors combined mean that some flowering plants around your veggie garden can reduce the amount of pesticides and work needed and hopefully get you that extra harvest.

So next time you plan a garden, literally think out of the box and consider planting your vegetables in unorthodox locations such as basil around your lavender or oregano as a ground-cover beneath your wattles.

Artificial nectar

If you don’t have room or otherwise cant plant flowers you could always try providing insects with artificial nectar honey solution, the recipe my lab uses is as follows:

Ingredients:

  • 180ml Honey
  • 1800ml Hot water
  • 10.8g Ascorbic acid
  • 1.8g Sorbic acid
  • 1.8g Paraben
  • 10ml 70% Ethanol

Dissolve the Sorbic acid and Paraben in the Ethanol.
Dissolve the Honey in the Hot water and allow to cool.
Add the Ascorbic acid to the honey and water and mix well.
Add the dissolved sorbic acid and paraben and mix well.
Store at -4oC.

Let me know if you try this!

Posted in Uncategorized | Leave a comment

A Problem of Nomenclature

Entomologists are in a unique position whereby they are often expected to be taxonomists as well as their field of specialty. However at the same time correct taxonomy can be of little use when communicating with the wider public.

When I first visited a vineyard for my PhD I realised the difficulty of communicating my research to the wider public. Talking about hymenoptera can result in blank stares, while talking about wasps may result in concerned comments about the European wasps (Vespula germanica) you have apparently found.

This example highlights how concerns can be inadvertently created by not taking the time to properly tailor communications to the target audience.
One way to better communicate research is to add descriptive prefixes to family groups that can help demonstrate the point that is being made. For example I have made it a habit to now to refer to parasitoid hymenoptera as beneficial wasps or native beneficial wasps.

Other issues to look out for are:

  • “bugs” meaning hemipterans not insects in general
  • “Insects” being interpreted as including spider, worms and all the other creepy crawlies
  • Parasitoids hosts: I have read concerned comments online about people worried that bio-control releases could parasitise humans!

Have any readers encountered similar issues?

DSC02553a native cockroach

Posted in Uncategorized | Tagged , | 1 Comment