Research+Paper+5

=**Conducting primary research**=

As you search for an answer to your **research question**, you will need to determine what kind of research will best fit your **rhetorical situation**. Will you collect data on your own (**primary research**), will you need to read the research results of others (**secondary research**), or will you need to do both? **Secondary research** is likely more familiar to you, but you shouldn't avoid **primary research** just because you have never conducted an interview or designed a survey. Sometimes **primary research** is essential for answering a **research question**.

Researchers often use more than one method of gathering data in order to avoid inaccuracies in the research and to confirm findings. Validating the research results is the researcher's responsibility, and you could confirm your research conclusions by finding two or three reliable sources or methods that provide similar results.


 * **Types of primary research** || **Method of data collection** ||
 * Observations || Gathering data through your own senses ||
 * Interviews || Asking questions of one or more people in person ||
 * Surveys || Asking questions of larger groups of people ||

=**Observations**=

When researchers conduct observations, they use all of their senses to note everything they can about a subject and its environment. Observations can be a useful way to gather firsthand information about a subject by relying on senses and note-taking skills. Observing a subject directly is sometimes a useful way to narrow your **research question** during the **invention stage**, to determine the extent of a problem or issue that you **research question** addresses, or to answer the question itself.

For example:


 * **how doe bilingual speakers switch back and forth between English and Spanish in casual conversation?** In this case, a researcher would observe the conversation of bilingual speakers and not when they switch back and forth between languages.


 * **At what times of day are voting locations most busy on election days?** To answer this question, a researcher might observe voting locations at various times of the day.


 * **To what degree does "light pollution" obscure the view of the night sky in a particular metropolitan area?** Part of the response to this research question should include an observation of the night sky on various occasions.


 * Including observation in your research plan**

Your **rhetorical situation** may or may not call for **observation**. Consider the following questions as you determine whether observation would be a useful data-gathering technique for your project.


 * **Is there a person, place, activity, or ritual that you might watch, hear, experience in order to answer your research question?** If so, you might want to directly observe your subject.


 * **Would observing your subject over time provide data that would help you answer your research question?** In this case, you might schedule observations at certain intervals that would help you collect the appropriate data.


 * **Would your presence as an observer influence the subject that you are observing?** This influence is often called the **observer's paradox** - the presence of the observer sometimes affects the environment being observed and, therefore, the data collected. Sometimes it cannot be avoided, but it is important for the researcher to be aware of his or her potential influence on the subject being observed.


 * **Could you participate in an event, ritual, or environment that would give you unique insight in the the subject you are studying?** If so, you might consider being a **participant observer**, an ethnographic term for a person who is collecting data and researching a subject while participating in or with it. In such a case, you need to be clear about your relationship to the subject when you describe your data collection method, and consider your perspective as a participant in your analysis of the data.


 * Conducting an observation**

Once you have decided to conduct an observation, what should you do? First determine where and when you will observe and have a plan for taking clar notes about your subject as you observe. Some researchers keep nots in a journal, and many organize their journals as **double-entry journals** to keep their observation and analysis/response separate. In a **double-entry journal**, one side is generally reserved for the observations themselves (which might even be organized by senses or themes) and the other side is used for interpretation, response, or comment on the observations.

Careful note-taking is essential to completing a useful observation, especially if you are the sole person responsible for data collection. Ideally, you will conduct your observation and then have time to return to your notes later, when you will have a different perspective. You might write down some of your interpretations as you conduct the observation (or immediately after), but then you might add interpretations in the fight-hand column when you read your notes later. As you read the notes, especially when you have some distance from the subject, patterns will emerge that might lead to interesting conclusions about your subject.

In addition to having a clear plan for taking notes during your observation, you should keep in mind that it is not possible to be entirely objective when you observe. Your goal, rather, should be to remain as neutral as possible in your observations. Your experiences, beliefs, and perspective will influence you, drawing your attention to certain sights, sounds, and smells (among other things). You should strive to experience as much as possible, though, and to be award of - and try to limit - the influences that could **bias** your observation. One way to do so is to have a method for conducting the observation. If you have never done an observation, try the following exercise to practice observation skills, and then consider using the strategy for collecting data to answer your **research question** as well.


 * Practice observing your subject **

Choose a space to observe for ten minutes. It might be a location in school. at work, or in your home. Make sure that you won't be interrupted for the ten minutes that you are observing, and bring a double-entry journal with you to record your observations. Note: You might practice this method by observing a space with which you are familiar before beginning your observation for your research project.


 * Start by observing the space immediately surrounding you, only going out one or two feet from where you are sitting. First, observe what you see, then what you hear, then what you smell, touch, and taste.


 * Extend your observation by a few feet. Again, record what you observe with each of your five senses.


 * Finally, focus on each sense and write down everything you observe - as far as you can see, hear, smell, touch, or taste.

Once you have finished noting your observations, take a moment to write comments on your observations in the other column of your double-entry journal. What patterns do you see? What surprised you?

=**Interviews**=

For some research projects, the researcher needs to ask questions of someone directly involved with an issue. Like observations, interviews might be used when you are trying to determine how to focus your **research question**, and they might also be used once you have defined your **research question** and want to answer it. Interviews are often integral to answering a **research question** as in the following examples.


 * **What are the writing practices of individuals in the software engineering field?** In this case, the researcher might interview one or more professionals in the field to find out what kinds of writing they do.


 * **How do international students use social networking applications to connect with friends and family?** A **research question** such as this might include an **interview** to hear the student's perspective, as well as an **observation** to learn how the student uses social networks for this purpose.


 * **How might climate change impact the flora of the eastern coast of North Carolina, and how soon would that change be evident?** For this project, the researcher might want to **interview** one or more experts on the subject. An observation wouldn't be feasible for a short-term research project since the **research question** is projecting into the future. If the researcher could conduct an extended project, though, observations might also help answer this question.


 * Including interviews in your research plan**

You might **interview** an expert in a specific field in order to gather data that will help you answer your **research question**. You might also interview people who are participants in or members of a community (such as a profession) that you would like to study. Depending on your **research question**, you will need to determine how many people to interview. As you consider whether interviewing might be a good method for your research, consider the following questions.


 * Would talking to experts in a particular field help you to answer your **research question**?


 * Would speaking with participants in an activity or members of a specific community help you to answer your **research question**?


 * What could you learn from speaking with someone that you can't learn from reading published information or from conducting an **observation**?


 * Could you conduct and **interview** on the phone or online if it is not possible to conduct the **interview** in person?


 * Conducting an interview**

If, after answering the previous questions, you have determined that you should conduct an **interview** for your research project, then you will need to start planning your **interview**. First you need to do three things:


 * **Set the interview time and location with the person (or people) you will be interviewing**. Try to select a location that will make the interviewee feel comfortable but not distracted, and keep the length of the interview reasonable. Thirty minutes to an hour is a reasonable length of time. Keep in mind that your interviewee is doing you a favor by setting time aside for you, so be careful to schedule around the interviewee's commitments. If you realize you need more time, ask to schedule a follow-up interview. Sending a confirmation before the interview and a thank-you note afterward is a nice, courteous touch.
 * **Write your interview questions**. If you are conducting an interview, you should come prepared with specific questions so that the interview is efficient and you get the information you need to answer your **research question**. Remember that the interviewee is giving you his time, so be prepared and professional. As you write your questions, avoid closed questions that elicit short yes/no responses (i.e., "Are professional in your field expected to write much?"). Instead, write open-ended questions that will invite the interviewee to talk about a particular subject. Open-ended questions generally start with //who, what, when, where, why,// or //how//, such as "What kinds of writing are professionals in your field expected to write?" or "How much writing is expected of professionals in your field?" In addition, make sure that you keep the number of questions reasonable for the time allotted. You might have a few questions in reserve in case you have additional time, and you might want to order your questions based on importance in case you run out of time.
 * **Decide how to record the interview**. You might want to digitally record the **interview** or you might take notes. When you choose your method, consider how important it will be to have the interviewee's exact words. Note-taking, as done in an **observation**, could work if you can just paraphrase and summarize the responses of your interviewee. An audio recording will work if you need the exact words of your subject, and a video recording might be best if it is important for you to notice expressions or gestures. Keep in mind that each method of recording (note-taking, audio recording, and video recording) could be increasingly distracting to your interviewee, and that will likely affect his responses. If you don't need a video recording, for example, then opt for a less distracting method. And if you do need a video or audio recording, make sure that your recording device has enough memory and a power supply to record the entire interview. Of course, in order to do a video recording, you would need to interview the subject in person (or have the subject use a webcam if your subject is agreeable). And for any recording, be sure to obtain permission from your subject before proceeding.


 * Draft interview questions **

Before you begin to write questions for your interview, write your responses to the following prompts.


 * What do you need to know from your interviewee? Why do you want to interview this particular person? Freewrite for five to ten minutes.


 * Once you have finished your freewrite, read your response. Look for patterns. Are there specific things that you need to know? Generate a list.


 * Now go through your list and prioritize the things you would like to know.


 * Finally, write a question for each item on your list; keep them open-ended. If you have several closed questions (questions that will elicit only a yes/no response), try inserting "Why?" at the end of such questions.

Once you have finished a preliminary list of questions, consider how many questions you will have time to ask. Are there some questions you should combine? Are there some that should be separated? Then try your interview questions out on a friend. Revise as you ask them, looking for questions that are awkward or confusing.


 * **Interview Questions** || **Responses** || **Comments/Analysis** ||

=**Surveys**=

Some research papers require asking questions of a large group of people and would not be manageable by doing individual interviews. In such cases, surveys are often the best option. Because a **survey** can take a considerable amount of time to design, distribute, and assess, it works best when the researcher has a well-defined **research question** that clearly calls for the kind of data a **survey** would produce - that is, data from a large group of people that could help to identify a widespread trend or phenomenon. **Surveys** can provide essential data for answering some **research questions**, as in the following examples.


 * How do American college students rate speakers of different dialects of American English on multiple characteristics such as intelligence, likeability, and friendliness?


 * How many citizens in my town generally vote on election days, and what reasons do they give?


 * How well do graduating seniors from my college feel the required first-year writing class(es) prepared them for the writing they did in other college classes?


 * Including a survey in your research plan**

To decide if a survey would fit the criteria for your **rhetorical situation**, consider not only the kind of data you would need to answer your **research question** but also how feasible it would be to conduct a **survey**. **Surveys** can take a tremendous amount of time to do well, and you would need access to the appropriate people who would be willing to complete your **survey**. As you determine whether or not you will conduct a survey, consider the following questions.


 * **What can I learn from a survey that I can't learn from an interview or observation?** **Surveys** enable a researcher to ask questions of a large group of people and also have the advantage of providing the researcher with the exact responses of the participants in text form.
 * **How many people would I like to poll in order to answer my research question?** If the number is larger than you could reasonably interview, a survey might be the way to go. The number for your research project must be small enough to be manageable but large enough so that you will have the interpretive power necessary to answer your **research question**. Many students who are designing **surveys** for the first time make the mistake of surveying an unrepresentative group.
 * **Do I have time to design a successful survey, distribute it, collect responses, and analyze the data?** **Surveys** can be quite time-consuming, and you will want to consider how much time you have to devote to your project as you design your **survey**.


 * Conducting a survey**

If you have determined that you should conduct a survey for your research project, in addition to choosing a representative group of participants, you need to pay close attention to two issues involved in design:

1) writing successful **survey** questions and

2) distributing and collecting your **survey**

First, as you write the questions for your **survey**, think about how many responses you will be analyzing. In other words, keep your project manageable, It is also critical to consider the kinds of questions and responses that will help you answer your **research question**. Two kinds of questions are most common in **surveys**: closed, or fixed-response, questions, and open-ended questions. Take a look at the different kinds of questions in the table below:

possible responses, so they are easily comparable with the responses from other participants. || Which item are you most likely to recycle? a. Cans b. Bottles c. Paper d. I don't recycle
 * **Type of Question** || **Issues to Consider** || **Sample Questions** ||
 * **Closed Question**
 * Multiple choice
 * True/false
 * Yes/No
 * Rating on a Likert scale || Gives a certain set of

I regularly place used soda cans in the recycling bin. ___True__ _False

Do you have access to the Internet at home? ___Yes__ _No

How confident do you feel in anticipating the needs of different types of readers when you write? 1 - Not confident 2 - Somewhat confident 3 - Confident 4 - Highly confident || more information, but responses are not as easily compared as answers to closed questions. || What technological support do you receive as a student?
 * **Open-Ended Question** || Allows participants to give

What differences do you perceive between taking a course online and taking a course in person? ||

Many surveys incorporate both closed and open-ended questions so that the researcher has statistics that can be compared but participants can also explain their answers, giving the researcher interesting information that helps to interpret the results of the survey.


 * Draft survey questions **

Before you begin to write specific questions for your survey, write your responses to the following prompts.


 * Who will you survey? How and why did you choose this group of people?


 * What do you need to know from your survey participants? Why do you want to survey them?

Read your responses to these questions. Look for patterns. Are there specific things that you need to know? Generate a list. Now go through your list and prioritize the thins you would like to know. Next, consider the kind of data that would be most useful in responding to your research question. Should some questions be closed questions in order to collect quantitative data? Which ones should be open-ended? Finally, write a question for each item on your list.

**Is the survey valid and reliable?**

Before using your survey with a large group of people, test it on a few friends. Look for points of confusion in the survey. Which questions should be revised for clarity?

As you test your survey, you'll also be determining whether the survey is valid and reliable. Ask the following questions as you revise the draft of your survey.


 * Is it **valid**? Does the survey actually assess what it's supposed to assess? For example, in the recycling scenario, asking "Do you think recycling is important? doesn't tell the researcher whether the person recycles. Also, if the questions are confusing, you might not be "testing" the same thing for everyone.


 * Is it **reliable**? Is the survey conducted in a similar way and scored identically for all participants? Your survey must be administered and scored consistently in order to be reliable. For example, participants should have a similar amount of time to respond, and if you are looking for patterns in open-ended responses, you should consider **inter-rate reliability**.


 * Distributing surveys and collecting responses**

Once you have designed your survey, you will need to plan how to distribute and collect responses. One of the biggest challenges in conducting a survey is getting people to respond, so you might consider some strategies that would help increase your response rate. All strategies should be disclosed when you report the results of your research.

=**Ethical considerations**=

If you are conducting research that will include other people (especially if you are asking questions of them), then you'll probably need approval from a group that oversees research that involves **human subjects**. Such a group is usually called an **institutional review board (IRB)**, and it exists to make sure that research on humans is conducted ethically and doesn't violate anyone's rights (especially privacy) or have adverse effects in health or well-being.

=**Interpretation of data**=

It is best to collect data from a variety of sources and in a variety of manners to compare responses and interpret the results. Researchers call this **triangulation** of the data. Consider this scenario: imagine that you conduct a survey and receive several similar responses to one question. But one of your participants gives a response that is completely opposite to all the other responses.

Once you collect your data, search for patterns that will help you draw conclusions. There are two major categories of data - quantitative and qualitative - and your analysis and interpretation will vary depending on which type of data you have collected.


 * Analyzing quantitative data**


 * Quantitative data** can be analyzed numerically. You can tally results and compare averages, look for statistical patterns, and determine whether a majority of respondents said the same thing. Of the types of research discussed so far, only **surveys** can generally be used to collect **quantitative data**. **Quantitative data** are often valued because they are more generalizable; that is, they are collected from a large population.


 * Analyzing qualitative data**


 * Qualitative data** are not intended to be analyzed statistically in the same way **quantitative data** are, but they can provide insight into a subject. Interviews and observations provide **qualitative data**, as do the **open-ended** responses on **surveys**. When you analyze **qualitative data**, you look for patterns and trends. In some cases, you might be searching for the number of times you observed a certain phenomenon or the number of times a participant used a certain word or phrase, but because of the small number of participants you would not draw statistical conclusions about the significance of a phenomenon or generalize the results to a larger population. In this case, you would want to carefully aggregate the data and have a second person also aggregate the data, identifying the occurrences of a phenomenon. This process is called **coding**. The second coder would help to provide inter-rater reliability. In other words, the results are more powerful because the researcher is not the only one verifying the patterns in the data. You can check your interpretation of the data with another person.

Regardless of whether you collect **quantitative** or **qualitative data**, strive to phrase your conclusion in a way that is appropriate to what is called the **interpretive power** of your study. You'll want to provide your readers with a clear understanding of the limits of extending your results, based on the scope of your study and the way you define your **rhetorical situation**. You could also use hedging terms to provide for the possibility that another researcher would find a different results since you will be unable to observe all situations or interview/survey all possible participants. Hedging terms include words like //generally, usually, often, might//, and //could//.


 * Presenting the results of primary research**

//**IMRAD Format**// When you present the results of primary research in a written format in many disciplines, especially those in the sciences and social sciences, your audience might expect you to follow what is often called **IMRAD** format. **IMRAD** is an acronym that stands for the following steps:


 * **Introduction** - generally an introduction to the subject, an overview of the secondary research on the topic, and a clearly stated research question. This section should answer "What's the subject and why is it important?"


 * **Methodology** - a clear description of the methods used to conduct the research, including a description of the participants, how they were selected, what was observed, what questions were asked, how data were collected, and how they were analyzed (i.e., if you had a second **coder** look at the data, this is where you would explain your process). This section should answer "What did you do?"


 * **Results** - a presentation of the data collected. This is generally separate from any analysis of those data, which comes in the next section. This section should answer "What did you find?"


 * **Analysis** - conclusions you have drawn from the data you collected. This section should answer "What do the results mean?"


 * **Discussion** - implications of the results of your research or suggestions you are making based on the results and analysis of your data. This section should answer "What are the implications of the results?" or, more informally, "So what?"

If you're not sure what format to use to present the results of your research, **IMRAD** provides a useful outline to follow.

//**Presenting Data Visually**// While **IMRAD** format provides an arrangement pattern for the written presentation of your primary research, you might also be interested in presenting the results of your data visually. A graphic depiction of data can often communicate the impact and importance of results more effectively and efficiently than a textual description can. Many researchers incorporate visual representations of data into written reports or oral presentations. Three of the most popular ways of presenting data are **tables, charts/graphs**, and **maps**. Depending on the nature of the data you have collected and the results you would like to highlight, you might find one of these methods more appealing and persuasive than the others:


 * //**Tables**// provide, in essence, a graphic organizer of data collected through primary research. They are especially sueful for presneting the results of **qualitative data**, since **tables** are not restricted to representation of numerical data. If you wanted to display common responses to survey questions, for example. a **table** would be a good way to do so.


 * //**Charts/Graphs**// (bar chart, line chart, pie chart) are used to represent numerical data and to show the relationships between data. Three of the most common types of **carts** are bar charts, line charts, and pie charts. A bar chart can be used to compare results between groups. Line charts are generally used to show a progression of a trend, perhaps over time. Pie charts often show distributions of responses to a particular question.


 * //**Maps**// are useful ways to present results if you have collected data that have geographical significance


 * What does your research plan look like now? **

At this point, take some time to plan the kinds of research you need to conduct. You might have already started a research plan that sorted though the different kinds of secondary resources for your project. If you are interested in conducting primary research, you can take that information and put it into a more comprehensive research plan, using the chart below.

You also need to decide when you are going to complete your research. Keep in mind that it will be of the utmost importance to manage your time well if you are conducting primary research; be sure to plan enough time to collect and analyze data.


 * **Types of Resources** || **Location of Resources** || **Deadlines** ||
 * **Primary Resources** ||  ||   ||
 * **Secondary Resources** ||  ||   ||
 * **Secondary Resources** ||  ||   ||
 * **Secondary Resources** ||  ||   ||
 * **Secondary Resources** ||  ||   ||
 * **Secondary Resources** ||  ||   ||