Observation as a Research Tool

One of the most valuable qualities of a researcher stems from their own worldview. How the researcher sees the world leads to the papers they right, the research they conduct, and the theories they hope to test. How they see the world makes their views and observations totally unique. No two people will have the exact same worldview and that leads to a diverse and constantly changing academic world. Observation is a wonderful tool, but difficult to use because of how unique every worldview is.

Observation is almost impossible to separate from your own personal biases. This is actually not a large problem, as there are easy ways to circumvent this. Since everyone has to face the same problems of bias in observation it is expected. Let me provide an example of differences in observation:

A young boy is given three options for a snack. He is given an apple, a marshmallow, and a bag of chips. The boy looks at the bag of chips and holds them in his hand before picking up the marshmallow and looking at it with obvious desire in his eyes. He seems hesitant and looks around before sighing and putting it down. The boy then picks up the apple and, with less desire in his eyes, chooses the apple for his snack. He later asks if he could have the marshmallow as well.

Depending on your own personal experiences, what you observe, and the reasons behind the boys actions, will be different than another researcher. One researcher might decide that the boy chose the apple because his parents always taught him that he must finish healthy food before he could have a treat. Another researcher might observe a boy who feels like he might get in trouble if he chooses the sugary snack, so he chooses the healthy one and hopes he can have the sugary snack later. A third researcher might see a child who knows that an apple is more filling as a snack and, even though the marshmallow is more of a treat, chooses the apple because he is more famished. He still tries to get the marshmallow after he is filled up, though.

Or maybe he was just traumatized.

Again, all these different observations occur because each researcher has had a different life experience and worldview. A researcher that grew up under parents who made them eat healthy food before a dessert would see one thing while a researcher who had very “hands-off” parents might see it differently.

Bias is unavoidable, but as academic researchers, we must do our best to address it and make it known. There are some biases that you will never be able to remove, but if you address it yourself, you are acknowledging it and therefore it can’t be used as an argument to discredit your paper.

The average scholarly debate.

So how do you try to remove bias from your own observations? The easiest way is to have multiple researchers observe the same occurrence and compare their observations. This is a simple way to remove some bias and hone in on the facts. Think of this method like a Venn Diagram. You all observe a participant and use your research to find the common factors that everyone saw. Another way is simply taking a step back and attempting to look at at things from a different perspective. This is handy outside of just research (looking at you ex-girlfriends). Doing this won’t come through specifically in your research, but will allow you to readjust your mind set before you decide your official observation.

Research is difficult and bias makes things even more difficult to present your research in a manner that will stand up to academic scrutiny and analysis.


Surveys: What Can They Bring To Your Research?

Arguably, one of the easiest methods of research (besides observation) is the survey. Design a survey, get it approved, and hand it out to the masses! This is a quick and fast way to gain a lot of research. With the introduction of the internet and technology, sending out surveys became ten times easier and a thousand times more effective. Doesn’t it? With the internet, you only need to share a link, and depending on how you share it, you can receive thousands of responses. This should make the survey a more valuable tool to the researcher, right?

Unfortunately, there are some downsides to the survey. In fact, there are many.

The first problem comes when you build your survey and how you word your questions. You MUST make sure that your questions don’t promote a bias in their wording. What do I mean? Well, let me provide an example:

     Question 1 – How would you improve the flavor of Mighty Moose Chocolate Ice Cream?

Question 2 – What would you like to see added to Mighty Moose Chocolate Ice Cream?

The wording in Question 1 implies that Mighty Moose’s Chocolate Ice Cream’s flavor needs improving, or that it is less than par. Just by this simple wording, you make your audience assume that the flavor is of a lesser quality. This is actually a big problem when it comes to creating a survey. How do other researchers know that you didn’t phrase your question like this purposefully to gain specific answers to support your view? With just a slight change, look at the wording in Question 2. This question doesn’t assume that the flavor of Mighty Moose Chocolate Ice Cream is good or bad, instead the question appeals to the personal taste of the audience taking the survey. If I was conducting a survey in order to determine if people disliked Mighty Moose Chocolate Ice Cream, I could phrase all my questions like Question 1 and probably have a lot of research to support my claim. But that research isn’t fully unbiased.

The second problem comes with authenticity in surveys. When you conduct surveys, you can only assume that the people who take your survey are being honest. Due to the anonymity of a survey, there is no way for you to check the accuracy behind the survey answers. This doubles up if they survey is taken around other people. Even though the survey is anonymous, the surroundings of a person taking the survey can influence their answers. For example:

     A survey is given to a class of 9th grade boys asking about their sexual experiences and alcohol consumption. Not every one of these kids may have participated in either of these options, but one friend sitting too close to another can influence the answer of a survey taker. Even though the answers might be anonymous to the researcher, these boys could still change their answers if not provided a safe place to take the survey.

A good way to prevent this unwanted influence is to provide a more private place for the survey to be taken. It could be a booth provided by the researcher that allows a participant to fill the survey in private or it could be as simple as asking the participants to spread out in a room so they have more privacy. Obviously, the first option would be more effective, but that might not be a possibility depending on your time constraint or budget.

A final problem with surveys are outliers. Outliers are pieces of data that are far away from the norm. For example, if you were to present a survey to the general population about how many times that go skydiving a year and a skydiving instructor happens to take the survey. Well he goes skydiving two hundred times a year, which is much more than the average person. His results alone could skew the average data. So, since the skydiving instructor is such an outlier, you decide to remove his answer. The problem comes with deciding who is an outlier and how many you can cut. It is possible to show bias just off of what data you decide to show and not show.

He skewed your survey with that goofy smile.

Surveys are a great resource for researchers and should be taken advantage of. In order to have strong research from a survey, you must create the survey very carefully and attempt to keep it unbiased.