Research Methods: Variables, Scales, Ethics, and Validity

Conceptualizing Variables: Age and Responsibility

If building a research study around the question "What is the relationship between age and responsibility?" one would have to take that question and conceptualize the elements that would lead to measurable variables. The subjects of the question — age and responsibility — are what we want to measure, and so we build around those two elements a study approach that will yield the most efficient data for measurement, or the operational data.

For this question, the operational data on age versus responsibility would best begin with establishing the age range of the participants to be studied. We could graph that range — for example, 13 years of age to 21 years of age. We could use as the research question: "After school each day, homework is (a) a priority, or (b) not a priority." Other conceptualized elements that could yield operational data might include the importance of getting to class on time, personal hygiene, exercise, diet, video game time, and television time. All of these data elements are conceptualized based on the original question and can serve as operational data points for measurement.

Taking one of the variables from the previous essay and constructing a scale around it yields useful information for analysis, comparison, and drawing conclusions about age versus responsibility. A scale applied to the data element intended to measure video game time might look like the following.

Measurement Scales: Likert, Thurstone, and Guttman

A 5-point Likert Scale measuring video game behavior could include items such as:

On weekdays I play video games: 1 hour per day — 2 hours per day — 3 hours per day — More — Less

On weekends I play video games: 1 hour per day — 2 hours per day — 3 hours per day — More — Less

I play video games alone: 1 hour per day — 2 hours per day — 3 hours per day — More — Less

I play video games with friends online: 1 hour per day — 2 hours per day — 3 hours per day — More — Less

Ethical Considerations in Human Research

I play video games: Before doing homework — After homework — While doing homework

The Likert Scale relies on devising as many input criteria around the objective as possible. In the example above, it is designed to elicit both the amount of time spent gaming and the relative importance of homework versus video games in the student's daily life.

A Thurstone Scale on the same topic might present the following agree/disagree items:

1. I like video games. — Agree / Disagree
2. I play video games every day. — Agree / Disagree
3. I play video games online. — Agree / Disagree
4. I play video games with friends online. — Agree / Disagree
5. I play video games after school. — Agree / Disagree
6. I play video games before doing homework. — Agree / Disagree
7. I play video games while doing homework. — Agree / Disagree

Based on the seven criteria above, a Guttman Scale would record each respondent's agree or disagree response across all seven items in a matrix format, with respondents listed in rows and items in columns. The key to Guttman scaling lies in the analysis: the table reveals which respondents agree or disagree with each criterion, and those percentages are then used to make comparative points about the criteria and about the broader relationship between age and responsibility.

When conducting research with human participants, a number of ethical considerations must be adhered to in order to protect participants' rights and safety. Institutional Review Boards (IRBs) and Research Ethics Boards review all proposed research at academic institutions to ensure that participants will be treated fairly and ethically. Some of the main tenets of ethical research include voluntary participation, safety from harm, confidentiality, and reporting guidelines.

Voluntary participation refers to the principle that participants in a study are not coerced or compelled in any way to take part. Participants must choose to participate freely and must feel free to withdraw at any point without consequence. Ensuring voluntary participation can sometimes affect the kinds of compensation provided. For example, if you are conducting a study on low-income families, you may want to pay participants; however, it is important not to pay them too much, as excessive compensation could create a sense of coercion — the participant may feel they have little real choice because they cannot pass up the financial gain.

Participants in research studies also have a reasonable expectation that they will not come to any serious form of harm. Any study involving potential harms or risks must clearly outline these possibilities at the outset, when the participant completes the Informed Consent form. In general, studies should avoid methodologies that could lead to harm unless a significant potential benefit outweighs the negative consequence. Medical research for terminal illnesses is a clear example: drug side effects are often severe, but those harmful effects are considered a justifiable risk when weighed against the potential benefit of finding a cure.

Confidentiality is a critical factor in research, and its degree of importance varies with the nature of the study. When studying a sensitive subject — such as sexuality or experiences of abuse — researchers must ensure that participants have a genuine sense of confidentiality and the opportunity to participate anonymously. Even when the subject is not especially sensitive, confidentiality must still be maintained: participants' information must not be shared with third parties or individuals not directly involved in the study. Because participants often share considerable personal information, they need assurance that researchers can be trusted. Ensuring confidentiality has many practical implications for methodology. Researchers conducting online surveys must verify that the website uses a secure connection and that data transfer is encrypted to prevent third-party interception. Once data is collected, it must be stored securely, and participants should ultimately be identified only by a code that cannot be linked back to their identifying details.

Although maintaining confidentiality is a primary obligation, extenuating circumstances sometimes require it to be broken. This typically falls under guidelines for reporting, by which researchers are required by law to report instances of child abuse or the potential for serious harm to self or others. If, during a study, a child participant indicates that he or she is being abused, the researcher would be required by law to report the abuse to the proper authorities, and could no longer keep that participant's data confidential. This obligation makes research on child abuse particularly difficult, yet such research is critically important. Parents are unlikely to consent to their children participating in a study if they believe it could lead to having their children removed from their care. One way researchers navigate this tension is by designing studies that ensure total anonymity of the data — so that not even the researcher can identify the source of a particular response — thereby allowing the researcher to circumvent the mandatory reporting requirement while still collecting valuable data.