2  Research Proposal

“A problem well stated is a problem half-solved.” (Charles F. Kettering)

Before beginning a research study, creating a proper research plan is critical. Working as a group, you will want to thoughtfully and thoroughly answer the following questions about your project in as much detail as possible. It is important that each group fully consider and create their research battle plan prior to beginning the work of deploying things (e.g., before writing survey questions). This will ensure that each group understands the motivations and purpose of each component of the project, and will hopefully encourage creativity.

Note

Not all questions below will apply to all projects. Think critically while preparing your research proposal. If you have questions about whether or not to include a component, ask Dr. Van Horn in advance.

To complete the research proposal:

  1. Create a shared document for your group to edit.
  2. Re-create the sections below, in order, in your document.
  3. Respond fully to each prompt in the order they are presented below, including your answers in their corresponding sections.
  4. Once complete, turn in the fill-out document in PDF format.
Important

This proposal requires you to complete all planning of your project, including finding pre-existing questionnaires that you will borrow, detailed breakdowns of measurements you will create yourself, etc. Once approved, deploying this research proposal should be as simple as following a recipe to make a cake.

2.1 The Big Picture

Before worrying about the minutiae, such as writing individual survey questions, you must understand your overall goals. Begin here.

  1. What is your research question?
Note
  • This is not a hypothesis. It is a more general statement of interest. Here are some examples:
    • Does TikTok use lead to negative mental health outcomes?
    • Are today’s youth truly “digital natives” that have greater understanding of computers than their predecessors?
    • Can you really be your true self online, or do conformity pressures make you hide aspects of who you are?
  1. What existing observations and research motivates your research question?
Note
  • Summarize some relevant discoveries from your literature review that made you think of your research question.
  • The goal is to avoid inventing your question “out of thin air.” Instead, use your literature review for inspiration for your research question.
  1. What are your hypotheses?
Note
  • It’s a good idea to have a few, as this will make your project bigger and easier to write about for your final paper.
  • Hypotheses are far more specific than research questions.
  • Hypotheses are testable and are constructed using independent and dependent variables. Here are some examples:
    • TikTok users will have higher rates of depressive symptoms than non-TikTok users.
      • IV: TikTok users vs non-users
      • DV: rate of depressive symptoms
    • 18-22 year olds will score lower on a test measuring computer know-how than 30-34 year olds.
      • IV: Younger vs older individuals
      • DV: computer know-how
    • Individuals will report higher rates of conformity on social media as compared to in-person social events.
      • IV: Social media events vs in-person events
      • DV: rate of conformity

2.2 Variables

  1. What are your (quasi)independent variables?
Note

Independent variables are the components in a hypothesis test that you manipulate. In an experimental design, these are directly manipulated, such as the dose of a drug.

In non-experimental designs, one uses quasi-independent variables, which are found pre-existing. For example, smokers vs. non-smokers, or TikTok users vs. TikTok non-users.

  1. What are your dependent variables?
Note

Dependent variables depend upon independent variables, and are the measured outcomes associated with any associated predictions arising from your independent variables. Dependent variables are chosen because it is believed that the independent variables have an effect on them. For example, if smoker vs non-smoker was the independent variable, reasonable dependent variables one could measure would include:

  • Cancer incidence
  • Blood pressure
  • Number/frequency of stroke

2.3 Measuring Variables

How will you measure your independent and dependent variables? Provide a detailed explanation for each variable mentioned in the previous section.

For some variables, measuring is trivial. For example, if you had the independent variable “TikTok user,” you could plan to ask participants, “Do you use TikTok?,” with two possible responses Yes/No.

For other variables (both independent and dependent), measurements can be far trickier to obtain. Consider the following common psychological variables that are easy to understand, but difficult to measure properly:

  • Intelligence
  • Problem-solving ability
  • Impulsiveness
  • Anxiety
  • Fear of missing out (FOMO)
  • Personality traits, such as openness

Measuring these kinds of variables may require several steps:

  1. Create or borrow a psychological construct that estimates your variable of interest.
  2. Define the variable operationally
  3. Create or borrow a set of questions that carefully measures your variable. Follow the advice found in the project manual here. When possible, use (and cite) an existing, published method for measuring variables. For example, Cohen’s Perceived Stress Scale (PSS-10) could be borrowed (and cited) to measure your participants’ subjective stress levels. In other words, make sure someone else hasn’t already done the hard work for you before you attempt to reinvent the wheel.

2.4 Participants

  1. How many participants will you need? If you’re doing a survey, the answer is typically “as many as possible.”
  2. Will they have any special qualifications/prerequisites? For example, do they need to be users of social media? Of TikTok in particular?
  3. How will you sample from your desired population(s)?
Note
  • If your population of interest is just “young adults,” you can simply use Capital’s research participant pool to obtain participants.
  • If you have a more specific population that cannot be reliably obtained through the participant pool, such as “college athletes,” you need to specify how you plan to gain access to individuals from that population.
  • Some populations are inherently difficult to access. For example, individuals that are concerned about their privacy might not want to respond to a survey that asks questions about their privacy behaviors.
    • Avoid targeting populations that are protected, such as minors, prison inmates, etc.

2.5 Planned Comparisons

Given your hypotheses and variables provided above, describe all comparisons you plan to conduct. Keep in mind that with a large number of survey questions, there are many possible comparisons that can be done. Some will be useful to your goals, and others will not. Broadly speaking, there are two types of comparisons: (1) comparisons used to test hypotheses, and (2) comparisons used for exploratory analyses. Please fully describe all of your planned comparisons for each type.

2.5.1 Comparisons used to test hypotheses

These comparisons are the most important, and involve comparing outcomes on a dependent variable across one or more independent variables. An example:

  • We plan to compare the mean self-reported anxiety scores for TikTok users vs non-TikTok users. This will involve separating our results into two groups, those of TikTok users versus those of non-TikTok users using our “Do you use TikTok” independent variable question. Then, an anxiety score will be calculated for each participant by summing together their individual responses on the 10 item self-perceived anxiety assessment. That total value will be an estimate of their overall anxiety. We will then calculate the mean anxiety score for each group (TikTok users vs non-users).
Note

Please create a bullet item for each comparison that you plan to conduct within this section of your research proposal.

2.5.2 Comparisons used for exploratory analyses

After describing how you plan to test your hypotheses in the previous section, you will likely find that you have other possible comparisons. Sticking with the previous example, your original plan may have been to combine the 10-item self-perceived anxiety assessment into a single anxiety score for use in testing your TikTok user hypothesis. However, it would also be possible to do other exploratory things with that same data, such as seeing which individual items in the assessment had higher scores. For example, if those 10 questions each asked about a different source of anxiety, you could inspect those results to see which source is the most common in your sample. This is very useful when writing about your results, as this added information might provide deeper insights into why your hypothesis test comparisons went the way they did.

Note

There will likely be many such possible comparisons. Please create a bullet item for each comparison that you plan to conduct within this section of your research proposal.