IES: Lesson - Experiments 📖

By diligently following the scientific method, we ensure that our environmental research adheres to the highest scientific standards. 

Experimental Design

The experimental part of the scientific method is often considered the most fun and exciting part of the scientific method. This is where you get to test your hypothesis and put all your previous research into action. But, before you can conduct your experiment, you need to design a robust experiment.

 

•  Control 
•  Groups 
  
• Replication

Control

You want to carefully plan your experiment, making sure that it is a controlled experiment. This means that you'll control all variables except the independent variable. This will ensure that your experiment is set up in a way that allows you to isolate the impact of the independent variable on the dependent variable.

First, you want to determine your independent variable and dependent variable. Remember, the independent variable is BEFORE the comma in your hypothesis, and the dependent variable is AFTER the comma in your hypothesis.

The things that we keep the same in our experiment are the control factors. Once you've identified your dependent and independent variables and planned the experimental groups, you will need to make a list of all the factors you want to keep constant during your experiment. For example, in our fertilizer experiment, you'd want to keep the following constant between and among each group:

• container size and shape
• amount of soil
• amount and frequency of water
• temperature
• amount and frequency of sunlight
• type of plant in each pot
• time at which you collect your data
• measurement technique
• how often you collect your data

 

Groups

Now, you will need to plan the groups in your experiment.

Think about how many different treatments you would like to have. For our fertilizer example, we might test 0 ml, 10 ml, and 20 ml of fertilizer.

The 10 and 20 ml groups are your experimental groups, because they are exposed to the independent variable.

The 0 ml group is your control, which is not exposed to the independent variable. This allows you to have a baseline for comparison.

For example, if all three groups grew similarly sized plants, you might be able to infer that the fertilizer treatments had little to no effect on the growth of the plants in your experiment.

 

 

 

 

 

 

 

 

Replication

If you only have one group for each treatment, then if something happens to that one group, you have lost that data.

For example, while working on my master's thesis, I was sampling corals. I put all my samples for a certain group in test tubes and on the way to the spectrometer, I tripped and all my samples for that group were lost when they spilled. If I'd had multiple groups (and not tried to analyze all the samples at the same time), I would have been able to use my replicates as backup.

If you have several groups for each treatment, you have a larger sample size, and your data will be more accurate. Having more than one group for each treatment is known as replication and improves the validity of your results and also serves as a back-up if something happens to one of the members of that group.

Typically, you should have at least 3 replicates, or a sample size of at least 3 for each treatment, as shown in the image above.

 

 

Now you are ready to set up your controlled experiment and start collecting data!

You should always have a control group, which is not exposed to the independent variable, in your experiment to serve as a baseline for comparison.

Your experiment should have at least 3 samples per treatment.

Before setting up your experiment, decide which factors you will keep the same for all groups, so you can be as sure as possible that any changes in the dependent variable are due to the independent variable.

Data Collection

During your experiments, you will collect data, or detailed recordings of experimental observations. This means that you will record everything – date, amount of water, temperature, amount of fertilizer for each group, and height of each plant. You also might write down your observations, such as the color of the plants or anything else you notice during your experiment. You want to have as much data as possible. You can always decide after the experiment is over that you don’t need some of the data you collected, but you can never go back in time and collect more data if you decide after your experiment is over that you need more. For our fertilizer example, we might decide that we want to collect the following data:

• temperature
• plant height
• amount of water
• date and time that each measurement is made

Data are detailed recordings of experimental observations. Collect as much data as you think you might need - you can always decide you don't need the data later, but you can never go back in time and collect more data.

Accuracy and Precision

Your data should be both accurate and precise.

Avoiding Bias

When collecting data, it's very importance to steer clear of bias and ensure randomness. Bias is characterized as a "systematic error introduced into sampling or testing by favoring one outcome or answer over others." In the realm of science, it's all too easy for our personal preferences to influence our choices because we may unconsciously (or consciously) desire a specific outcome. Maintaining randomness in data collection is an important method for preventing bias. However, it's equally important to remain vigilant for potential bias factors when collecting data or designing experiments.

As you record data during your experiment, it's important to do so impartially, or without any bias. In the realm of science, researchers carry a moral responsibility to conduct experiments without bias. Unlike disciplines such as music or literature, where emotions and feelings are often intentionally evoked, science operates in an objective, unemotional realm. While scientists can be deeply passionate about their subjects and may excel when they have genuine enthusiasm, they must refrain from allowing personal emotions to obscure their experiments and the process of data collection.

Scientists have a moral and ethical obligation to keep their personal biases and feelings out of their experiments and data collection.

All experiments require a control group to eliminate other factors that may interfere with the experiment. The control group represents what happens without experimental interference or modification.

Work through the interactive below to "collect" all of the Design an Experiment FRQ tips:

[CC BY-NC-SA 4.0 Links to an external site.] UNLESS OTHERWISE NOTED | IMAGES: LICENSED AND USED ACCORDING TO TERMS OF SUBSCRIPTION - INTENDED ONLY FOR USE WITHIN LESSON.