Week 10: Introduction

By Shant Shahoian, lightly edited and updated by Adina Lerner

 

Introduction

This week we study bias, advanced Google searches, and research orientation. While these seem to be a hodgepodge of unrelated items, you'll find that discussing them together makes quite a lot of sense at this stage in the course. You will have access to the prompt for the Final Research Project, which should put much of what we've done so far into perspective. Also, understanding this week's lecture will be integral to your group's approach to this project.

Bias

Much of the reading this week will cover how language exposes bias. That is, we can have positive or negative descriptors, often for the same idea. Words with similar denotations can have different connotations. 

For example, you fall in love with Sam, who does not like to spend the money they have. Sam takes you for long walks on dates, gives you flowers they picked from a nearby neighbor's house, and takes you to restaurants, like MacDonalds. Sam may have a lot of money or they may be poor. All we know is that Sam does not like to spend money. You love Sam, so you describe them as frugal, economic, fiscally prudent, or even thrifty. Later, Sam breaks your heart, so you call them cheap, penny-pinching, or stingy. That is all of the words in bold here mean the same thing. They denote the same idea: the quality of being reserved with one's spending. But the words we use can include positive or negative connotations that impact how people interpret the idea. These connotations can also manifest as positive or negative descriptors, as you will see in this unit.

Beyond diction, bias is present when the author of the source has ulterior motives, when something can be sold based on the information provided, or when the piece contains biased information itself. We discussed how to approach these issues by using lateral reading and fact-checking resources. But bias can also manifest in a variety of other ways.

  • Inclusion or Omission: When you tell people you want to impress about Sam, you may include all the good they have done well but exclude negatives. Or the opposite could be true. That is, bias is in the author, the words, and the decisions of what to include. You want your family to love Sam, so you do not mention their penchant for running over cats for fun. This does not put Sam in a good light, so you exclude it. Was it pertinent to the story you told your parents about Sam? Perhaps. Is that a lie? It depends on what you promised. If you say, "Here is all I want to tell you about Sam," then leaving out the bit about their thirst for felicide is fine. But generally, if your parents want to know about what kind of person Sam is, excluding this sociopathic behavior is problematic. You're omitting something relevant. In other words, just looking at a source isn't enough to determine if it is biased. There are so many levels.
  • Even-Handedness: Are both sides of the issue presented?
  • Proof: Is there any evidence or examples to support such a statement? Or is it just asserted?
  • Generalizations: Does the source make broad generalizations or stereotypes? For example, does the author/source use statements like everyone knows or all sensible people…Is using stereotypes a legitimate way to make an argument? What impression do you as an information consumer have towards someone who uses stereotypes to make claims?

Another very real dynamic is the publishing source of the content. Examine this bias chart below. The vertical axis refers to the degree of factual reporting and the horizontal axis speaks to the extent to which sources are left-wing (progressive) or right-wing (conservative). The dotted green box indicates the most reliable sources, yellow outlines the resources with mixed reliability, orange is for mostly unreliable sources, and red is for unreliable sources. You'll notice that there are reliable sources on both sides of the political spectrum, but clearly, if a source is hyper-partisan, they can't be unbiased. The Wall Street Journal is just as biased as NPR, but both do a lot of factual reporting. But it's important to note that even with factual reporting, there is some bias. Think for a moment what this means. No matter how neutral we wish to be, we're human and bound to see the world through our own lenses. Neutrality is a goal but not a destination. We're not robots. On the other hand, Rachel Maddow on the left and Sean Hannity on the right are not journalists at all.  They fashion themselves "commentators" who tell you what to think instead of reporting as journalists. 

 

Advanced Google Searching

You'll want to be aware of this chart as you conduct research. Perhaps your topic requires a lot of contemporary sources, sources that are still early in the information cycle. Do you want to take information from InfoWars ("The Lizard People are coming!") or Occupy Democrats (I mean, c'mon, it's in the name!). You can omit certain sources, as you'll find out. You should also be careful that the keywords you use are, in fact, as neutral as possible. Be cognizant of your own cognitive biases. Links to an external site.

 

Deductive and Inductive Reasoning

Six Degrees of Kevin Bacon Links to an external site.

Did you ever play this party game called "Six Degrees of Kevin Bacon"? It's popular among film fanatics. Essentially, the goal is to name any actor to the player, and the player has to connect that actor to Kevin Bacon (who was a popular actor in the 1990s) in six steps. If the player cannot connect Bacon to the named actor, they lose. If they can, they win. It goes something like this:

    1. Six Degrees of Kevin Bacon: Doug Rand *.
    2. Well, Doug Rand was in the movie Arthur and the Revenge of Maltazard with Lou Reed, and
    3. Lou Reed was in the film One Trick Pony with Blair Brown,
    4.  And Blair Brown was in the movie Loverboy with Kevin Bacon. Done. Three degrees!

*Prof. Lerner's cousin Links to an external site.. Zendaya is two degrees. 

Adept players pick actors who they think the player will not be able to connect to Kevin Bacon, but the important point for us is that the players try to connect two disparate people. They use links to create a chain from the stated actor and Kevin Bacon. When you research you should create a chain from the quoted evidence to the point you want to make.

Deductive Reasoning  

Disclaimer: This is a gross oversimplification of complex logical structures and is intended to provide you with a brief overview of general approaches to research.

You should understand the basics of deductive reasoning -- and its cousin, inductive reasoning -- in order to understand the structure of research. (We will not discuss abductive reasoning in this class, but it's also one approach to a research paper.)  This will make sense in the very next section of this lecture.  Deduction is “the mental process of moving from one statement (‘All human beings are mortal’) through another (‘Socrates is a human being’) to yet a further statement (‘Socrates is mortal’)” (Barnet and Bedau 73).  Generally, we move from general ideas to a specific conclusion that is necessarily true. There are many deductively valid forms, with fancy names like modus ponens and modus tollens, but we will focus on the generic syllogism. This syllogism has a very simple and familiar form:

Example of Deductive Syllogism
Deductive Syllogism Example

A = B
B = C
Therefore, A = C
  1. Khlono is a turtle.
  2. All turtles have shells.
  3. Khlono has a shell.

You may remember the mathematical equivalent of this: the transitive property. 

For a deductive syllogism to be sound, the syllogism must be both true and valid. Are each of the general assumptions below true? If not, the syllogism is not true and therefore cannot be sound – but it can still be valid. Are the ideas linked together validly; do these two general assumptions lead logically to the conclusion? If so, the syllogism is valid, even if it is not true.

That is, the form follows the transitive property outlined above, but if the assumptions therein are not true, the syllogism is not sound or necessarily true. It does not work to prove the conclusion F=JC. Let's look at another one.

  • The whale is a large fish.
  • All large fish have scales.
  • Therefore, whales have scales.

What do you think? Well, whales are not fish; they are mammals. All large fish do not have scales. Because these are untrue, we know the conclusion (Whales have scales) is untrue. But again, the form is valid:

  • W=LF
  • LF=S
  • W=S

For the syllogism to be sound, that is necessarily true, the terms must be true and the syllogism must be valid.

Could it be the other way around? Could the assumptions of a syllogism be true but the syllogism not valid? Sure!

  • John breathes air.
  •  All humans breathe air.
  • Therefore, John is a human.

Let's say it's true that John breathes air. It's certainly true that all humans breathe air. And, it's entirely possible that John is a human. But even if the terms are true, the form is not valid. That is, John could be a chipmunk. A chipmunk breathes air, and all humans breathe air -- but that does not mean that chipmunks are human!

Remember, deductive reasoning is the logical equivalent of the mathematical principle called the transitive property.

  • A = B
  • B = C
  • Therefore, A = C

Induction

While deduction, if valid and true, produces a conclusion that is sound and, therefore, undeniably, 100% true, induction uses “observed cases to reach a conclusion about unobserved cases” (Barnet and Bedau 78).  Generally, in induction, we move from specific ideas to a general conclusion that is probably true. The more specific ideas we include, the more likely the conclusion is true -- but it will never approach the kind of 100% truth we get with a perfectly sound deductive syllogism.

There are many kinds of inductive reasoning, such as induction by enumeration and analogical induction, but we will stick with plain old general induction. Look at this inductive argument. Here, we notice a variety of suggestive details that lead us to a general truth that is probably true.

  •   Khennef reaches for his toothbrush.
  •   Khennef grabs the toothbrush.
  •   He holds the toothbrush with the bristles facing up.
  •   He reaches for the toothpaste.
  •   Conclusion: Khennef is about to brush his teeth.

Is it possible that Khennef is just going to fake us out, like my son did when he was younger? He would get ready, put the brush up to his mouth, and then run away. Yes, it is possible, but the more details we notice about Khennef, the more we are convinced of the likely truth that he is brushing his teeth. 

We use inductive reasoning every day. We notice someone with a clipboard at a market, they have a monogrammed polo shirt, and they walk toward you. You have probably induced that they are going to ask you to sign something or join a cause. 

  • Thghl puts a giant beach ball in his car.
  • Thghl places a portable cooler in the trunk.
  • Thghl is wearing Speedos, flip-flops, and shades.
  • Thghl puts sun block on his nose.
  • Conclusion: Thghl is going to the beach.

Again, could Thghl just do all this and drive around Glendale? Sure! It's possible but not probable. The more details we include, the more convincing it is that Thghl is going to the beach -- but no matter how much we include, it will never be 100%. It will just become increasingly probable.

Research Orientation

Let's be honest. If you're like most people, research means asking the easiest question to answer then finding the sources to "prove" the answer you've already chosen. Research indicates that this is true of everyone, that most people want the easiest source to find, whether or not it is accurate. I mean, does every student care if the quotes they have for their research paper are true? It's often a mad dash to the finish line. Most research papers, to use a hospital analogy, are more triage than cooperative care. Students feel the need to "finish" more than the need to "get it right." In this class, there is no research paper to submit, so we can spend time focusing on the actual research process.

Research is not linear, not a straight line from idea to finished product. It's a mercurial and dynamic process, sometimes involving a few steps back in order to move, maybe diagonally, forward. Reread that last sentence. It's a doozy. And it's honest. Research is philosophically dialogic, each stage reiterative, each move a conversation with the step before, all while requiring you to keep asking yourself questions and refining your expectations as you make steady progress. 

Real research is, in a nutshell, messy.

Right now, each group has a tentative research question. If this question is, indeed, a research (and not a regular) question, you'll use search strategies, your awareness of bias, and a host of other skills you learned in this class to find excellent resources to answer that question. If you find answers right away, you didn't really have a research question, and you'll return to the drawing board. As our reading Links to an external site.and videos (1)  (2) about the research process indicate, the creation of a real research question takes time, and finding a good question is more process than a declaration.

A good research question challenges you to conduct research in an area that is not fully understood yet. Or, in some cases with professional researchers, a good research question is one that can be answered sequentially.

Let's say your question is in a standard form: "Does the impact of A on B make C more likely?"

You may find a few sources on "the impact of A on B and the relationship between this and C." That is, you find some sources that say the impact of A on B makes C less likely and some that say C is more likely -- or that it has no impact on C. You start conducting your research.

  • You find a bunch of sources that show the impact of A on B does make C more likely but you find no controversy. This is not a research question. This is a regular question. (Ex. Does producing more gas-powered cars cause more fuel consumption making global warming more likely? Of course.)
  • You find a bunch of sources that show the impact of A on B does make C more likely but you find other experts who contest this. This is a research question. (Ex. Cars with lithium batteries reduce carbon emissions thereby making the rapid rise in global warming less likely. It's debatable.) 

In this case, you use inductive reasoning, which is probably the more common method in undergrad research papers:

  • You itemize a list of sources that all contend to show the impact of A on B does make C more likely.
    •  The impact of A on B makes C more likely in way V 
    • The impact of A on B makes C more likely in way W 
    • The impact of A on B makes C more likely in way X
    • The impact of A on B makes C more likely in way Y
    • The impact of A on B makes C more likely in way Z
    • If V, W, X, Y, and Z are true, it is likely that the impact of A on B does make C more likely. (Conclusion based on inductive reasoning) Each new bit of evidence makes the conclusion more likely.
      • You note the disagreements with this conclusion.
        • You uncover these disagreements by using other reliable sources.
          • These disagreements make you realize that in way P, the impact of A on B definitely does not make C more likely. This runs counter to your findings about ways V-Z.
          • What do you do? Omit pertinent information?
            • No.
          • Modify your question? Maybe.
            • You rephrase the question: "Does the impact of A on B make C more likely in cases of Alpha" (assuming way P is NOT Alpha).
          • Write a more temperate thesis later? Maybe.
            • "The impact of A on B makes C more likely, except in the way P."

What does the deductive equivalent look like?

But some of the most exciting research is when nobody has quite uncovered the relationship you are exploring.

  • You find many theorists that explain the impact of A on B but none of them seem to mention C.
  • But the impact of A on B produces some quality X. 
  • Then you find, in separate articles, that X does make C more likely.
  • You just contributed a new idea to the field of research by combining information from other sources in an original way that nobody else has by using deductive reasoning.

Mathematically, it would look like this.

A ---> B

A---> B = X

X ^ C

Therefore, A--->B ^ C

Silly Example:

  • A vegan diet leads to lower triglycerides. (Dietary journal)
  • Vegans since birth have 0 triglycerides. (Dietary journal)
  • Newly discovered Y rays are harmful to humans in cutting-edge ion particle accelerator research. (Physics Journal)
  • Y rays latch onto triglycerides in order to harm the body. (Physics journal)
    • Therefore, there is reason to believe that vegans since birth will have higher resistance to Y rays.

This was a long way of explaining that research could be a compilation of debated articles that support an argument or a combination of sources that, logically, reveal a new connection. You may find many articles that support chemotherapy and radiation for patients with stage 4 pancreatic cancer and some who disagree with this approach because it is statistically unlikely to help and may exacerbate the problem. You can then conduct research on the many arguments for this treatment and investigate the limitations of each approach, eventually coming to a complex thesis: Patients under the age of 35 with Stage 4 pancreatic cancer should undergo chemotherapy but not radiation if they have no underlying medical condition or auto-immune disorders.

However, the person who cures pancreatic cancer won't be compiling what others have already done. They will be combining the work of others, often with original experiments of their own, to come up with new information which they will contribute to the ongoing scholarly conversation.

 

Disclaimer: This is an oversimplification of the many ways research can be conducted. I would recommend Thomas Mann's Library Research Models Links to an external site. for a comprehensive take on research orientation models.