Useful Reading
Campbell, Biology 6th Ed - pgs 16-20
Campbell, Biology 7th Ed - pgs 19-24
Vocabulary
Hypothesis - a working assumption, a possible cause, an educated guess, a tentative answer to some question. Contains only a statement of possible fact.
Prediction – what you expect to be true if your hypothesis is correct, formed as an if-then statement.
Dependent variable – in an experiment, the variable that will be measured or observed in response to the experimental conditions.
Independent variable – in an experiment, the variable that you will manipulate.
Control – in an experiment, the control group is a treated exactly the same as your manipulated group except that the manipulation is not performed
Science is a framework for gaining knowledge of the world. Science is usually applied by following the “scientific method” – a set of procedures and practices. While the scientific method is often described as an inflexible procedure, most scientists typically proceed with more flexibility and imagination. By using the scientific method, scientists aim to reduce bias and subjectivity and discover accurate, consistent, non-arbitrary representations about the natural world.
Science relies on available evidence, and is often corrected or improved if better evidence becomes available. This self-correction keeps science honest and objective.
Several websites provide more discussion on the scientific method:
http://phyun5.ucr.edu/~wudka/Physics7/Notes_www/node5.html
http://teacher.nsrl.rochester.edu/phy_labs/AppendixE/AppendixE.html
How to apply the scientific method
There are several steps to the scientific method:
1. Observe some aspect of the universe and formulate a question about your observation.
2. Invent a tentative description to explain what you observed, called a hypothesis, that is consistent with what you have observed.
3. Use the hypothesis to make predictions.
4. Test those predictions by experiments, further observations, or synthesizing other available data. Modify the hypothesis in the light of your results.
5. Repeat steps 3 and 4 until there are no discrepancies between theory and experiment and/or observation.
Step 1: Observations and Questions
Making observations and questions about the natural world can be more difficult than you might imagine. We make innumerable observations every day about the world around us, but we often don’t think to question these observations. This is what scientists must do. Sometimes the observations and questions arise during scientific investigation of a related question.
This step in the scientific method requires curiosity and good perception of the world.
Here, we will consider the question “What determines how big a radish plant is?” or “What limits growth in radish plants?”
Step 2: Proposing a Hypothesis
A hypothesis includes a suggested explanation of the observation you made. It will generally provide a causal explanation or propose some correlation.
Some hypotheses, however, are better than others.
· A hypothesis must be testable. If your explanation (hypothesis) includes variables that are difficult to quantify or measure, you will not be able to test your hypothesis.
· A hypothesis must be falsifiable. This means that you have to be able to disprove your hypothesis. In fact, when we test a hypothesis, we cannot conclude that our hypothesis is “true”; all we can conclude after our test is that our hypothesis is supported or wrong.
· Having multiple hypotheses is good – try to test one at a time.
· Restrict the hypotheses that you test to the “educated guesses”. While any testable, falsifiable hypothesis can be used in the scientific method, some may seem highly unlikely as explanations right from the beginning. You will save time if you start testing those that seem plausible given what you already know about the natural world and the system you are observing.
Proposing a hypothesis can be the most imaginative and insightful step of the scientific method. Sometimes the best explanation of your observations is not obvious. Think how long it took scientists to hypothesize that kids resemble their parents because information is encoded in genes in every cell!
For our radish example, we could propose the following hypotheses:
1) Light availability limits a plant’s growth
2) Water availability limits a plant’s growth
3) Nitrogen availability in the soil limits a plant’s growth
4) Competition from other radish plants limits a plant’s growth
5) Amount of love and affection from the horticulturalist limits a plant’s growth
6) God decides how big a plant will be
Based on what we know about plant biology, 1-4 seem like plausible hypotheses. They may, in fact, all have an effect on plant growth, and they may be interrelated. Hypothesis 5 is not a good hypothesis because “love” and “affection” are hard to quantify. Hypothesis 6 is not a scientific hypothesis because it is not falsifiable (how would you ever scientifically disprove that a supernatural being has an effect on plant growth?).
Null and alternative hypotheses:
The logic of traditional hypothesis testing requires that we set up two competing statements or hypotheses referred to as the null hypothesis and the alternative hypothesis. These hypotheses are mutually exclusive and exhaustive.
A null hypothesis, H0, is typically the hypothesis you are trying to disprove/reject during your research – it is a hypothesis of “no effect”.
The alternative hypothesis, H1 is the hypothesis to be accepted if / when the null is rejected. The null hypothesis is assumed to be true unless we find evidence to the contrary.
The final conclusion once the test has been carried out is always given in terms of the null hypothesis. We either "Reject H0 in favor of H1" or "Do not reject H0".
Since the alternative hypothesis is often the hypothesis that the researcher would like to be true, it is sometimes referred to as the Study Hypothesis or Research Hypothesis.
The hypotheses listed above for plant growth are all alternative hypotheses. There are null hypotheses that correspond to each. For example, the null hypothesis for (1) is “Light availability has no effect on growth in plants.” This is the hypothesis we set out to support or reject by designing our experiments and tests.
For more on null and alternative hypotheses, click here.
Step 3: Making predictions based on your hypothesis
Once you have a hypothesis, you want to figure out some way to test it and come up with predictions of what will happen in the test. Predictions are if-then statements: “if my hypothesis is true, then I expect to see X result in an experiment”, “if my hypothesis is true, then I expect further observations to show X”. Predictions are usually test-specific.
In our radish example, we can make the following predictions:
1) If light availability limits a plant’s growth, then shaded plants will grow less than plants in full sun.
2) If water availability limits a plant’s growth, plants given more water will grow more.
3) If competition from other radish plants limits a plant’s growth, then a radish grown next to other radishes will grow less than an isolated radish.
Step 4: Testing your hypothesis
Tests of your hypothesis can include:
1) Further observation
For example, if your hypothesis were “Birds sing to attract mates”, then you may make further observations on the timing of singing in birds. Your prediction may be “If my hypothesis is true, then birds will sing mostly during the beginning of the breeding season, and singing during other times of the year will be non-existent or greatly reduced”
2) Synthesis of data
For example, if your hypothesis were “The decline of the sea otter along Pacific shores has previously been and still is largely due to human hunting”, then you may look at historical records of sea otter distribution and abundance as well as fur trade records. Your predictions may be “If my hypothesis is true, then the initial decline in sea otter abundance should coincide temporally with the beginning of human hunting of otters and fur trade.” “If my hypothesis is true, then the declines would be heaviest in the most heavily hunted regions.”
3) Experiments:
Experiments are one of the most often used and frequently the best tests of you hypothesis. When designing an experiment, you want to think about the appropriate procedures, variables and controls. Here, we’ll use as an example the hypothesis “Light availability affects the growth of radish plants.”
Variables are the defined, measurable factors that may be important for your hypothesis. The dependent variable is the factor that you want to question and measure the response of in your test. The independent variable is the variable you are manipulating in the test. Other variables that may affect your dependent variable should be controlled variables – they should be identical for all groups. By controlling all other variables, we can ask whether it is the independent variable specifically which leads to a response in the dependent variable.
For our example,
What is the dependent variable?
Plant growth
What is the independent variable?
Light
What are the controlled variables?
Because we suspect other variables may affect plant growth, we will control Nitrogen, water, and plant density.
The Procedure of the experiment is the design and steps that you employ. To begin with, you want to choose one or several levels of treatment, establish a control group, and run several replications of each treatment/control.
For our example, we want to explore the effects of light
What levels of treatment should we use?
A good option would be to subject plants to varying degrees of shading (decreased light). We may use three levels of treatment: 25% light-reducing shading, 50% light-reducing shading, and 75% light-reducing shading.
What is our control group?
A plant near the shaded plants, but with no shading over it – full light
How should we replicate?
We may choose to place 15 plants in each group and average the growth results.
Based on this test, what would we predict if our hypothesis is true? Our prediction would be: If the hypothesis is true, plants under shading will show reduced growth compared to the plants in the control group. We might also predict that plant growth is less in each successive level of treatment, from 0% shading to 75% shading.
Step 5: Repeat the whole process
When you test your hypothesis, you may find that the hypothesis is not supported, or you may make additional observations that lead you to modify your hypothesis. This iterative nature is one of the keys to success when applying the scientific method.
You may also want to test several related hypotheses. For example, if your tests support the hypothesis that light affects growth in radish, you may want to expand your examination of plant growth by testing the other hypotheses we proposed. You may also test for interactions of variables (eg maybe light only affects growth when the plants have more than adequate water, but in water limitation, light has no effect on growth).
For more
on observations, hypotheses, and predictions:
http://biology.clc.uc.edu/courses/bio104/sci_meth.htm
http://en.wikipedia.org/wiki/Scientific_method
Review Questions
- What
is the difference between controlled variables and a control in an experiment?
- Define
hypothesis and prediction and describe how they differ.
- If you
wanted to perform an experiment to test the hypothesis “Birds sing to attract
mates”, how might you manipulate birds and what dependent variable(s) would you
record?
-
Imagine you have the hypothesis that “Productivity in vegetation (amount of
plant material produced per unit time) affects insect diversity in that
habitat”.
-What variables are important
to test you hypothesis?
-How would you use data
available or collect new data from established habitats?
-How would you sample
habitats?
1) sample only one habitat
2) sample multiple habitats of the same productivity
3) sample one habitat each of different productivities
4) sample multiple habitats each of different productivities
-What data would you collect
in these habitats?
-How would display the data (graph)?
-What would you predict your
graph to look like if your hypothesis is true?
- You
observe that sparrows are usually in broad-leaf trees while jays are usually
seen in Pinon Pines or Junipers. What hypothesis could you propose to explain
the difference in tree usage by these two types of birds?