Choosing a Type of Survey: Knowledge vs. Perception 1 Composing Inquirer Guide: Surveys Introduction: Choosing a Survey Type There are many ways to approach an assignment that requires surveying. However, with most student assignments, there is short time given to complete the requirements. Therefore, it is necessary to follow an efficient and effective process in order to produce concrete data from which the project can be completed. Where definitive data is gathered from a survey, it will be easier to organize and analyze the results. The underlying purpose of a survey assignment is to measure whether you are able to identify a proposition and to design and implement a plan to gather the data necessary to prove the proposition. Professors are usually looking to evaluate the approach taken, not necessarily the argument made, but also the ability to identify and acknowledge the limitations of your research. They will be looking to determine whether your questions directly correlate to your objective and are correctly structured to yield the desired results. The “Steps for Conducting a Survey” are described in Composing Inquiry and this supplement uses those steps to help you determine which type of survey would yield the best results for your assignment. There are many kinds of surveys that researchers conduct, but in this section we will examine the two most common types: perception and knowledge surveys. Perception surveys pose questions about what the subjects feel or think on a given topic. Knowledge surveys present questions to determine what the subjects think they “know” as factual information. The most critical difference in these two types of surveys is that subjects are asked to identify rather than evaluate. These guidelines are for students in need of completing a basic survey assignment and who have possibly never previously conducted a survey or encountered the particular difficulties of achieving an efficient and effective structure. *Note: Not all steps are covered in this supplement please refer to the main text for the complete list of “Steps for Conducting a Survey.” Determining the Research Questions Shown in Figure 5.11 below is a Sample Perception Survey. In this survey, the “Style Survey” group wanted to find out how much students care about their personal style and where their style influences originate. The Style Survey group first hypothesized that most students care about dressing mainstream and having the “latest and greatest” brand fashions to impress friends and demonstrate a wealth status. Therefore, their research question was “how much do students care about their personal style?” While the topic appeared to have a clear goal, the complexity of the subject slowly became apparent as the group attempted to compose a survey. In fact, an indefinite number of factors contribute to someone’s “personal style.” Therefore, the survey would have to entail an extensive list of questioning in order to pinpoint a subject’s beliefs. Plus, there is the additional dilemma posed by the fact that respondents may not provide truthful or consistent responses on personal-belief questions, which would skew the survey results. ©Kimberly St. Onge Margaret Marshall Composing Inquiry: Teachers’ Resources Prentice Hall 2008 2 The Composing Inquirer Conversely, the “Pop Culture Survey”, shown below in Figure 5.10, researched whether students know more about highly publicized celebrities or current political figures. This group hypothesized that political figures are not well known among college students, and therefore, put together a series of questions aimed at having their subjects identify (with pictures) the names of specific political figures. In a knowledge survey, there is a right and wrong answer. This gives the surveyors more certainty over the validity of the subject’s responses, since the answer depends upon factual knowledge rather than truthful perceptions. Therefore, the first crucial step to conducting a survey is to pick a research topic that will facilitate the development of objective “test” questions so that definitive data is gathered, rather than data that is conceptualized by the respondent. Defining Variables In this stage of the process, the details and specific word choice that will be used to achieve the goals of the survey are determined – a problematic task for the group undertaking the Style Survey. The process of forming questions that pinpoint a person’s “style” could involve an infinite number of variations. Reasonable people will differ on the meaning of personal “style,” which could refer to clothing, body art like tattoos and piercings, hair color, shape, or one’s overall body image. The group’s first question in the survey, “how important is having your own style?” leaves the respondent open to consider too many options. Where the researcher’s question lacks a clear concept of “style,” control is placed in the hands of the respondent to determine for themselves what they believe “style” encompasses. This creates the potential for a significant level of variability in the nature and meaning of the responses. This variability has a direct impact on the reliability of the results because if the researcher conceives style as one’s clothing and the respondent’s conception is focused on one’s overall body image, then the response has little value in terms of proving the researcher’s underlying proposition. Also, since the word “style” appears in many of the questions on the survey, this same effect could occur in more than once place. As a result, the collective data would be extremely flawed and results are based on different meanings of the term depending on how the respondent reads the question. The Style Survey also encountered an obstacle in terms of what constitutes caring about style? How does a person evaluate the level of importance that style has for them? What does it mean to “care”? Again, the definition of “care,” for example, is too broad for there to be a standardized understanding of the term and thus allows for multiple interpretations. In a perception survey, it is critical to define these variables because it is necessary to ensure that subjects read the question with a consistent meaning. For students conducting a survey for the first time, defining terms precisely can be difficult to accomplish. In the Pop-Culture Survey, the group did not need to spend as much time considering the definition of their variables because there simply were none. The variables were specified by the determination of which political figures and celebrities were most fair and appropriate to “test” their subjects on. This type of specification is more objective, more easily established within the time allotted for the assignment and within the ability of first-time surveyors. If the knowledge Choosing a Type of Survey: Knowledge vs. Perception 3 survey consisted of a research question like “how much do students know about cars?,” then the surveyor would need to determine what feature they want to focus on, for example car brands or models, engine sizes, or prices. This kind of variable definition is easily distinguishable and if done correctly the respondents will straightforwardly be able to understand what the surveyor is looking for. Developing questions for polling (interview protocol) or questionnaire With a knowledge survey, the researcher has control over the consistency and meaning of the responses, which makes the data more reliable because there is less room for multiple interpretations to take place. For example, in the Pop Culture Survey, the question “what is the name of the person pictured?” does not require one to define any key words because the response that is being requested is obvious to the respondent. The respondent will only consider the picture provided and nothing else. Another example of a knowledge question is giving a list of names and asking, for example, “who was the former Secretary of Defense?” also leaves the responder with no choice to consider options other than those provided as opposed to manufactured ideas not contemplated by the surveyor. By performing a knowledge survey, only a few questions are needed to generate the data necessary to support a supposition about the level of political versus pop-culture knowledge. Conversely, forming questions for a perception survey requires an entirely different approach. As with any survey, the answers available to the respondent should relate directly to the supposition the researcher is trying to prove. For example, in the Style Survey, Question 6, “Describe the way you dress by choosing two of the best options below”, with a list of different style labels given. The problem here is: how does knowing how a person dresses going to tell you about how much they care about their style? Someone that dresses punk may care just as much as someone that dresses preppy, but the only difference is their sense of style. Plus, it is difficult to put one’s style into a particular category since style changes with occasion, mood and surrounding culture. As a result, this question does not lead the researcher to any clear conclusions and therefore does not provide any value in supporting the underlying supposition of the researcher. Question 5, in Figure 5.11, highlights another problem with perception surveys. By asking a respondent “what influences you more when making judgments about other people?” the surveyor may get an answer that is less than truthful because the respondent fears appearing shallow with an answer that indicates that appearance plays the biggest role when making judgments. The respondent may feel that there is a “right” answer that they are supposed to choose instead of giving their true belief and such “right” answers distort the survey results. In Question 7, “how important is it that people around you (friends, family, peers) approve of your style?”, more than one question is being asked to the respondents It might be important for someone that their friends approve of their style, but not their family, which causes conflicted answers. Something to beware of when creating a survey, especially when asking about preferences or beliefs, is that question responses are not always congruent with each other depending on the variable, such as friends versus family. It is easy to put too many ideas into one ©Kimberly St. Onge Margaret Marshall Composing Inquiry: Teachers’ Resources Prentice Hall 2008 4 The Composing Inquirer question when you’re trying to keep the survey short. This is not an effective way to obtain good results and being simple and concise, as the knowledge survey demonstrates, is more important. For this reason, when assessing what someone feels or believes on a topic like style, it can be very difficult to extract the results needed because there are too many factors that contribute to such a topic. Analyzing Results The data collected from the perception survey is provided in Figure 5.12 and is organized by the number of times a respondent chose a specific answer. Questions 1 and 2 should have allowed respondents to answer based on a Likert Scale rather than choosing from the three levels of importance. The provided answer “somewhat important” was the most popular choice in both questions, but this does not actually tell the researcher just how important. The question remains unanswered and the data cannot be used to support any argument because it does not prove that style is important or is not important to students. In addition, the data appears scattered and unrepresentative of any obvious conclusions in all of the questions, which caused problems for the group as they approached the written portion of the assignment. These problems are a result of the flawed tactics used in questioning, which was presented in the previous sections. Also, problems arose which resulted from the lack of connection among the questions. The initial objective needed to be strengthened and variables more clearly defined. Still, it is undeniable that perception surveys require a more involved approach to surveying because of the depth of research and questioning required to make inferences about people’s beliefs. Alternatively, knowledge surveys generally produce clearer and more objective data results. Being able to determine whether a respondent answered correctly or incorrectly to a question will lead to better organization of data and supply obvious support for an argument. The ability to demonstrate and use the factual information gathered gives the researcher more authority during the writing process and leaves fewer questions for their audience. The limitations of a knowledge survey are based more on technicalities like sample size and range of questions rather than issues with interpretation which are usually the hardest limitations to explain. KNOWLEDGE VS. PERCEPTION SUMMARY DO… BEWARE OF… Create a knowledge survey Creating a perception survey Keep ideas basic, simple and concise Trying to encompass too many ideas at once with a survey Create a survey utilizing “test” questions Clearly understand the argument you want to Forming questions that respondents may fear answering truthfully make before putting together survey questions Broad terms like “style” Form questions directly relating to the argument proposed – what will the data show? Allowing room for interpretations in your survey questions – maintain Force your respondents to only consider the control! Not providing your questions with an adequate range of response options Choosing a Type of Survey: Knowledge vs. Perception options you provide for them Keep in mind survey limitations Figure 5.10 Sample Knowledge Survey Year______ Major______ Sex______ G.P.A._____ Name that Person: a. b. c. d. e. Oprah Winfrey Maya Angelou Condoleezza Rice Maxine Waters I don’t know a. b. c. d. e. Joshua B. Bolten Dick Cheney Robert M. Gates Tony Blair I don’t know 3. The president of Iran is: (Circle One) a. al-Maliki b. Ahmadinejad c. Talib Kweli d. Mahmoud Zahar e. I don’t know 4. France’s president is: (Circle One) Pro American Anti American I don’t know 5. Who was the former Secretary of Defense? (Circle One) a. Dick Cheney b. Donald Rumsfeld c. John McCain d. Rudolph Giuliani e. I don’t know 6. After the attack on Pearl Harbour, who said... "This is a date which will live in infamy” (Circle One) a. Andrew Jackson b. John F. Kennedy c. Franklin Delano Roosevelt d. Winston Churchill Name that person: a. b. c. d. e. Andre Agassi Rafael Nadal Roger Federer Rod Laver I don’t know a. b. c. d. e. Eric Dane Jason Thompson Patrick Dempsey Orlando Bloom I don’t know 9. How many children do Brad Pitt and Angelina Jolie have? (Adopted & Biological) (Circle One) 1, 2, 3, 4, 5, 6 Don’t know 10. Nicole Richie & Joel Madden are expecting... (Circle One) a. a ring b. a baby c. a divorce d. a cd together e. Don’t know 11. Who is the host of the 2007 VMA's (Circle One) a. T-Pain b. Jack Black c. Britney Spears d. Nelly Furtado e. Don’t know 12. Which of the following celebrities was recently arrested for DUI, while driving with a suspended license and had cocaine their car? (Circle One) a. Lindsay Lohan b.Leonardo DiCaprio b. Caroline Kennedy d. Britney Spears e. Don’t Know ©Kimberly St. Onge Margaret Marshall Composing Inquiry: Teachers’ Resources Prentice Hall 2008 5 6 The Composing Inquirer Figure 5.11 Sample Perception Survey Figure 5.12 Collected Data from Sample Perception Survey Check One: Male ____ Female ____ Age: ____ 1. QUESTION How important is having your own style? Circle One Extremely important Somewhat important FEMALES - 44 MALES - 31 TOTAL - 75 1 Extremely important Somewhat important Not important 17 24 4 11 16 4 28 40 8 2 Extremely important Somewhat important Not important Brands One of a kind 9 22 13 23 21 6 14 11 18 13 15 36 24 41 34 Latest and greatest Your own trends Style/appearance Personality 25 18 14 30 14 17 10 21 39 35 24 51 6 Preppy Punk Urban/Street Flashy Business/Sophisticate Self-created/artistic Gothic Metro Athletic Comfortable/casual Different style Grunge/hippy 20 1 3 4 6 1 0 0 7 23 6 1 9 0 3 0 5 0 0 1 14 15 4 1 29 1 6 4 11 1 0 1 21 38 10 2 7 Friends Family Religion Athletics/sports Fashion icons Music Art Hobbies/interests Job/career I don't know 17 9 0 4 10 3 2 7 2 10 10 2 1 5 1 2 0 5 0 9 27 11 1 9 11 5 2 12 2 19 8 Yes No Strongly agree Somewhat agree Disagree 37 7 13 27 4 27 4 15 16 0 64 11 28 43 4 Not important 2. How important is it that people around you (friends, family, peers) approve of your style? Extremely important Somewhat important Not important 3. What is more important to you? Circle one: brands or one of a kind If brands, name two that you like____________________________________________ 4. What is more important to you? Check one: ____ Latest and greatest/keeping up with the trends ____Starting your own trends and it doesn’t matter as long as it looks good 3 4 5. What influences you more when making judgments about other people? Circle one: Their style/appearance or Their personality 6. Describe the way you dress by choosing two of the best options below: ____Preppy ____Gothic ____Punk ____Metro ____Urban/Street wear ____Athletic ____Flashy ____Comfortable/Casual ____Business/Sophisticated ____Different style everyday ____Self-created/artistic ____Grundge/Hippy 7. What/who influences you the most to dress this way? (Pick Two) ____Friends ____Family ____Religion ____Athletics/Sports ____Fashion Icons/Celebrities *Please name celebrity:____________________ ____Music ____Art ____Hobbies/Interests *Please list hobby:______________________________ ____Job/Career *Please list job:______________________________________ ____I don’t know what influences me the most 8. Does your style reflect your personality? Circle One: YES or 9. Do you agree that personal style says a lot about someone? Circle one: Strongly Agree Somewhat Agree Disagree 5 NO 9