Applied Communication Research Methods A hands-on guide for applying research methods to common problems, issues, projects, and questions that communication practitioners deal with on a regular basis, this text demonstrates the relevance of research in professional roles and communication and media careers The second edition features updated material that covers major communication research methods including surveys, experiments, focus groups, and observation research while also providing key background information on ethics, validity, reliability, concept explication, statistical analysis, and other current topics It continues to foster student engagement with research through its numerous features and practical activities, including: • • • • Research in Depth—examples of methods as applied in scholarly research Reflect & React—problems and issues that promote reflection and discussion Voices from Industry—Q&As with professionals working in communication industries End-of-Unit Activities—exercises that reinforce concepts and content The text is ideally suited to both undergraduate and graduate courses in mass communication research methods Online resources, including sample syllabi, PowerPoint slides, and test banks are available on the companion website: www.routledge.com/cw/boyle Michael P Boyle is a Professor in the Department of Communication and Media at West Chester University, USA Mike Schmierbach is an Associate Professor in the Donald P Bellisario College of Communications at The Pennsylvania State University, USA Applied Communication Research Methods Getting Started as a Researcher Second Edition Michael P Boyle and Mike Schmierbach Second edition published 2020 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2020 Taylor & Francis The right of Michael P Boyle and Mike Schmierbach to be identified as authors of this work has been asserted by them in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988 All rights reserved No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe First edition published by Routledge 2015 Library of Congress Cataloging-in-Publication Data Names: Boyle, Michael P., 1975– author | Schmierbach, Mike, 1976– author Title: Applied communication research methods : getting started as a researcher / Michael P Boyle, Mike Schmierbach Description: Second edition | New York : Routledge, 2020 | Includes bibliographical references and index Identifiers: LCCN 2019040222 (print) | LCCN 2019040223 (ebook) | ISBN 9780367178710 (hardback) | ISBN 9780367178727 (paperback) | ISBN 9780429296444 (ebook) Subjects: LCSH: Communication—Research | Communication—Methodology Classification: LCC P91.3 B69 2020 (print) | LCC P91.3 (ebook) | DDC 302.2072/1—dc23 LC record available at https://lccn.loc.gov/2019040222 LC ebook record available at https://lccn.loc.gov/2019040223 ISBN: 978-0-367-17871-0 (hbk) ISBN: 978-0-367-17872-7 (pbk) ISBN: 978-0-429-29644-4 (ebk) Typeset in Sabon by Apex CoVantage, LLC Visit the companion website: www.routledge.com/cw/boyle Unit Overview Preface xxiii Basic Principles of Research and a Guide to Using this Book Basic Concepts of Research 12 Scholarly Research and the Creation of Knowledge 40 Ethical Research 80 Concept Explication and Measurement 100 Reliability and Validity 122 Effective Measurement 152 8 Sampling 186 Experiments and Threats to Validity 222 10 Survey Research 260 11 Content Analysis 286 12 Qualitative Research 306 13 Approaches to Qualitative Analysis 336 14 Descriptive Statistics 358 15 Principles of Inferential Statistics 386 16 Multivariate Inferential Statistics 426 Index455 Contents Prefacexxiii Basic Principles of Research and a Guide to Using This Book Principles of Research Empirical Systematic Intersubjective Cyclical and Self-Correcting Using This Book Steps to Success: The Research Process Voices From Industry: Jessica and Ziggy Zubric—Customer Experience Consultants 10 Suggested Readings for Further Exploration of Unit Topics 11 Examples 11 Advanced Discussion 11 Basic Concepts of Research Variable Unit of Analysis 15 Independent Variable 15 Dependent Variable 16 Third Variable 17 12 13 C o n t e n t s Relationship 18 Prediction 18 Positive Versus Negative Relationship 19 Mediation 20 Moderation 22 Reflect & React: Moderation 22 Causality 23 Time Order 24 Confounds 24 Theory Inductive Versus Deductive Reasoning 25 26 Reflect & React: Developing and Testing Theories 28 Falsification 28 Research in Depth: Theories in Communication 29 Parsimony 30 Voices from Industry: Meghan E Kennedy—Search Engine Marketing Strategist, Tower Marketing 31 Scope 32 Research Questions 33 Hypotheses 33 Activities 36 Activity 2A: Getting Started With a Research Question or Hypothesis 36 Activity 2B: Independent and Dependent Variables 37 Activity 2C: Causality 38 Suggested Readings for Further Exploration of Unit Topics 39 Examples 39 Advanced Discussion 39 Scholarly Research and the Creation of Knowledge 40 Knowledge 41 Bias42 Authority43 Research in Depth: The Power of Authority 44 Intuition 45 Reflect & React: Counterintuitive Findings 45 Tenacity46 viii C o n t e n t s Replication 46 Research in Depth: Understanding the Reproducibility Crisis 47 Triangulation 49 Voices from Industry: Chris Nietupski—Communication Consultant 50 Secondary Analysis Publication Bias Goals of Research 51 52 53 Steps to Success: Goals of Research 53 Application 54 Exploration 55 Description 56 Explanation 57 Qualitative Versus Quantitative 58 Parts of a Research Article 59 Steps to Success: Effectively Reading a Research Article 59 Abstract 60 Literature Review 61 Method Section 61 Results Section 62 Discussion Section 63 Future Directions 63 Reference List 64 Appendix (or Appendices) 64 Tables 65 Figures 66 Steps to Success: Identifying Research Articles 67 Other Types of Articles 68 Reflect & React: Effective Tables and Figures 69 Literature Review Article 69 Meta-analysis 70 Data-Driven Journalism 71 Activities 73 Activity 3A: Reviewing the Theoretical Literature 73 Activity 3B: Triangulation in Everyday Life 74 Activity 3C: Comparing Qualitative and Quantitative Research 75 ix m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s For example, suppose you want to predict whether someone would be willing to endorse a message on a social networking site, such as by clicking a “like” or “+1” button Ordinary regression would not be appropriate for such a test because the dependent variable is dichotomous But logistic regression could be used You might predict that evaluations of whether the content was enjoyable, credible, and interesting would all uniquely contribute to willingness to click the button By running the analyses, you might find significant positive coefficients for enjoyment and interest but no unique effect of credibility As with any regression analysis, this would show which variables have the clearest, most consistent relationship once you control for the influence of all other variables regression ® Unit 16 standardized coefficients ® Unit 16 independent variable ® Unit logistic regression ® Unit 16 variable ® Unit positive versus negative relationship ® Unit 446 Odds Ratio In ordinary least squares regression, we can evaluate the standardized coefficient to get a better sense of the relative influence of different independent variables by imposing a shared scale on those measures Logistic regression does not produce a standardized coefficient However, the analysis does produce an odds ratio, which can be used to compare the influence of different variables in a similar fashion and offers greater utility First, we must understand the idea of odds, which reflect the probability that something will occur divided by the probability it will not For example, a fair coin has odds of that it will come up heads There is a probability it will come up heads and a probability it will come up tails As 5/.5 = 1, the odds are 1, or perfectly even When the odds are greater than 1, we are saying the event is more likely to occur than not occur For example, the odds of seeing a presidential advertisement during an election are far greater than If 95 percent of voters are exposed to at least one ad, the odds that any given voter will see an ad would be 95/.05, or 19 On the other hand, when the odds are less than 1, we are saying the event is less likely to occur than not occur For example, the odds that a given individual will attend a political rally during an election are likely pretty small If only 5 percent of people go to a rally, then the odds would be 05/.95, or about 053 The odds ratio evaluates how these odds change as the value of a given independent variable changes For example, while the average person might have small odds of going to a rally, someone who has volunteered for the campaign might be much more likely to go If nonvolunteers have odds of 05 but volunteers have odds of (still less than even, but considerably higher), the odds ratio would be 10—that is, the individual is 10 times more likely to go if they volunteered than if they didn’t The odds ratio thus represents the amount you have to multiply the first set of odds by to get the second set for each one-unit increase in the IV Here, because the ratio is positive, it means that as the IV (volunteering) increases, the odds of the DV also increase This is a positive relationship On the other hand, consider the example of seeing m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s an advertisement Because ads tend to appear more on traditional media, we might expect that people who spend lots of time on social networking sites would be less likely to see an ad Suppose that for someone who spent 1 hour on social networking sites, the odds of seeing an ad are 19, but for someone who spent 2 hours, they are 9.5 The odds ratio would be 5—if you multiply 19 by 5, you get 9.5 We would further expect that when the hours increased from to 3, the odds would again drop—from 9.5 to 4.75 (9.5 × 5 = 4.75) This would be a negative relationship, as the odds ratio is less than The odds ratio ranges from virtually (for a very negative relationship) to essentially an infinitely high positive number (for a very positive relationship) Steps to Success: Selecting the Appropriate Statistical Test Box #1 — Start here: What level of measurement is your independent variable? Interval or ratio Go to Box #3 Nominal What level of measurement is your dependent variable? Do you have more than one IV (including control variables)? Nominal No Interval or ratio Chi-square test Yes Go to Box #2 No Are all your IVs and your DV dichotomous (only two values)? Stop: You cannot analyze these variables Select different measures or recode the variables Yes Logistic regression Figure 16.5 A flowchart for selecting statistical tests 447 m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s ANCOVA (interval/ratio variables as covariates) Figure 16.5 Continued 448 m u l t i v a r i a t e Nominal Box #3 — Start here: What level of measurement is your DV? Is your DV dichotomous (just two values)? Yes i n f e r e n t i a l s t a t i s t i c s Interval/ratio Do you have more than one IV? No No Stop: You cannot analyze these variables Select different measures or recode the variables Correlation (Pearson’s r) Yes Is your DV nominal? Yes No Regression Logistic regression Figure 16.5 Continued 449 m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s Activities Activity 16A: Preparing for Multivariate Statistical Analysis See terms: multivariate analysis; inferential statistics; level of measurement For this activity, develop a research question that you would like to answer that considers the relationship among three or more variables Perhaps there is a variable you think is important to consider as a control variable, or perhaps you would like to explore the relationship between multiple independent variables and a single dependent variable As you develop this research question, think about how you will measure the variables and the statistical tests that will be most appropriate for testing your relationship Consider the following questions: Which of the variables will be the dependent variable? Independent? Control? What level of measurement you expect your different variables to be? Based on these initial thoughts, what types of statistical analyses would you likely run to explore the relationships among the different variables? What makes this test (or tests) appropriate? What assumptions does this test have? How would you ensure that your data conform to those assumptions? 450 m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s Activity 16B: Critically Assessing Statistical Analysis See terms: statistical analysis; multivariate analysis; level of measurement For this activity, locate a research article that conducts some type of multivariate analysis Start by describing the nature of the relationship (or relationships) and the test (or tests) that are used to assess the relationship Use the following questions to guide your assessment: What particular tests are used in this research report? Identify each of the different types of variables (i.e., independent, dependent, and control) What is the level of measurement for each of those variables? Are these particular tests appropriate? Why or why not? Could the researchers have used different tests to test these relationships? How would you evaluate the quality as far as how the tests and their results are reported? Do the researchers clearly describe the results of their tests? Is there any information missing that would help your evaluation of the data and the results as described? Are the conclusions drawn from the data accurate based on the tests conducted? 451 m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s Activity 16C: Multivariate Statistics in Survey Research See terms: multivariate analysis; regression; inferential statistics For this activity, locate a research article that uses survey research as a methodology and some form of regression analysis as the analytical approach Use the following questions to guide your analysis of the article: 452 What kind of regression was used for the primary analysis? Was this type of analysis appropriate to the hypotheses or research questions asked? What are the main findings based on the analysis? What specific statistics the authors discuss in their report? In what ways they address statistical significance (or lack thereof)? m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s Activity 16D: Determining Appropriate Statistical Tests See terms: chi-square; t-test; regression; correlation; ANOVA; logistic regression For the following research scenarios, determine the appropriate statistical test you would use to test the relationships described In some cases, there may be multiple options that are appropriate Therefore, be sure to indicate a clear rationale why you have selected the test you did Finally, consider the assumptions of each test in making your decisions Research scenario Appropriate test You are interested in comparing test scores for the final exam in a research methods course for three different groups Group receives a weekly personal tutoring session beyond the regular course content Group receives enhanced course materials ahead of lecture Group receives both the weekly personal tutoring and the enhanced course materials You work for a political campaign and are interested in testing the relationship between age and the amount of money donated to the candidate for whom you work As part of your work for an advertising agency, you have been asked to explore the relationship between parenting and media preferences In particular, the agency would like to know whether there are significant differences between people who are parents of kids under 18 and those who are not for their favorite social media platform As part of your work for a nonprofit agency, you are interested in understanding whether the likeability of the agency’s spokesperson is related to the amount of money donated by individuals However, you expect that individual attitudes toward philanthropy and community volunteering also affect the amount of money donated Your director has asked you to control for those two variables in assessing the relationship between spokesperson likeability and money donated You are working with a client who is putting on a series of informational clinics on vaccine awareness The client is interested in testing a series of ratio- and interval-level independent variables including income, education, interest in health, and skepticism toward science on whether or not people come to the informational clinic (a “yes” or “no” answer) 453 m u l t i v a r i a t e i n f e r e n t i a l s t a t i s t i c s Suggested Readings for Further Exploration of Unit 16 Topics Examples Eveland, W. P., Jr., Hayes, A. F., Shah, D. V., & Kwak, N (2005) Understanding the relationship between communication and political knowledge: A model comparison approach using panel data Political Communication, 22(4), 423–446 doi:10.1080/10584 600500311345 Kahlor, L., & Eastin, M. S (2011) Television’s role in the culture of violence toward women: A study of television viewing and the cultivation of rape myth acceptance in the United States Journal of Broadcasting & Electronic Media, 55(2), 215–231 doi:10.1080/08838151 2011.566085 Advanced Discussion Hayes, A. F (2009) Beyond Baron and Kenny: Statistical mediation analysis in the new millennium Communication Monographs, 76(4), 408–420 doi:10.1080/03637750903310360 Hayes, A. F., Glynn, C. J., & Huge, M. E (2012) Cautions regarding the interpretation of regression coefficients and hypothesis tests in linear models with interactions Communication Methods and Measures, 6(1), 1–11 doi:10.1080/19312458.2012.651415 Hayes, A. F., & Matthes, J (2009) Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations Behavior Research Methods, 41(3), 924–936 doi:10.3758/BRM.41.3.924 Mundry, R., & Nunn, C. L (2009) Stepwise model fitting and statistical inference: Turning noise into signal pollution The American Naturalist, 173(1), 119–123 doi:10.1086/593303 Preacher, K. J., & Hayes, A. F (2008) Assessing mediation in communication research In A. F Hayes, M. D Slater, & L. B Snyder (Eds.), The Sage sourcebook of advanced data analysis methods for communication research (pp. 13–54) Thousand Oaks, CA: Sage Publications 454 Index Note: Page numbers in bold refer to the main entry for a listed term Pages in italic indicate a boxed term abstract 42, 59, 60–1 anchor 113, 114, 154, 166, 168, 178–79 ANCOVA (analysis of covariance) 427, 434, 435–6; see also ANOVA anonymity 82, 87, 88, 266, 268, 273; see also privacy ANOVA (analysis of variance) 403, 406, 427, 434–5, 438, 440–1; one-way 411–3 appendix 42, 62, 64–5, 66 application 42, 54–5, 56 artifact see social artifacts attribute 17 102, 109, 111, 236, 343, 407, 411 authority 6, 41, 43–4, 46, 155 bell curve see normal distribution between-subjects design 224, 227–8, 229, 233 bias 4, 41, 42–3, 58, 124, 126, 158, 161, 174–5, 203, 210–1, 214, 307, 308–9; interparticipant 224, 245; publication 48, 52–3, 70, 397–8, 432–3; researcher 87, 137, 158, 224, 237–8, 249, 398; selection 224, 227–8, 250; see also validity bivariate analysis 387, 405–6, 415, 427, 429, 431; see also ANOVA; chi-square; correlation (Pearson’s r); t-test Bonferroni correction 389, 400, 437; see also type I error case study 307, 309–10 causality 14, 16, 23–4, 56, 139, 223, 225, 227, 229, 233, 261–2, 265, 278, 287; confounds 14, 23, 24–5, 225, 229, 233, 276, 406, 436; time order 14, 23, 24, 25, 225–6; see also control variable1; relationship; spuriousness ceiling and floor effect 248; see also validity cell-phone-only household 195, 213, 269–71 census 187, 196, 197, 201, 202, 394 central tendency 359, 360–1, 365, 374; mean xxi, 62, 113, 201–2, 359, 361–2, 365–6, 368–9, 370–1, 374, 376, 377, 378, 379, 391, 393, 401, 404, 405, 409–10, 412, 443; median xxi, 359, 361, 362–3, 366, 374, 376, 377, 378, 379; mode xxi, 359, 364–5, 372, 373–4 chi-square 387, 389, 394, 406–9 clarity 141, 154, 176, 177 closed-ended 154, 160–1, 170–1, 165–6, 172, 179; see also open-ended cluster sampling 188, 196, 203, 209, 211–2; see also probability sampling coding 135, 170, 172, 237, 288–9, 289, 291, 293–4, 295, 297, 298, 299, 314, 340, 341–3, 350; guide 289–90, 298; sheet 290 coefficient: standardized 440, 443, 446; unstandardized 343, 442–3, 445 coerce 81, 84, 191–2 cohort 262, 276, 277; see also longitudinal design comparison group see treatment group complete observer see participant observation complete participant see participant observation completion rate 191, 268, 271, 274 computer-aided content analysis 288, 291 concept 58, 60, 101, 102–4, 105–6, 107–8, 128–36, 140–1, 141–2, 154, 169, 170, 171, 174, 176–8, 199, 262, 289, 298, 325, 337–8, 341, 343–4, 350, 357; see also concept explication concept explication 8, 61, 101–2, 105–6, 108, 123, 133, 135, 140, 289 conceptual definition 101–3, 104, 105, 106, 107–8, 133, 135, 136, 140, 176; see also operational definition I n d e x concurrent validity see validity confederate 82, 85, 88–9, 90–1 confidence interval 187, 190, 192, 198, 199–200, 202– 3, 209, 211–2, 263–4, 362, 369, 374, 387–8, 432–3; see also estimation; sampling distribution confidence level 198–9, 200, 202–3, 209, 263–4, 374, 433; see also estimation; sampling distribution confidentiality 81, 86, 87–8, 268; see also privacy confounds see causality consent see informed consent; form 83–4, 85, 86, 162 constant comparative technique 341, 343–4, 350; see also grounded theory construct validity see validity content analysis 71, 124, 160, 195–7, 214, 241, 287–8, 289, 291, 293–4, 295, 296, 309, 343, 348 content validity see validity contingency question see filter and contingency questions control group 223–4, 227, 233, 237, 238–9, 243, 246, 250 convenience sampling 188, 203, 204–7, 242; see also nonprobability sampling convergent validity see validity conversation analysis 338–40, 351 correlation (Pearson’s r) 134, 387, 395, 396, 401, 405–6, 413–15, 428, 439, 441–2, 444; see also bivariate analysis criterion validity see validity critical value see significance Cronbach’s alpha 125, 128, 141, 142–4, 144–5, 178 cross-sectional design 261, 262, 264–5, 276–8; see also longitudinal design cyclical and self-correcting 3, 5–7, 41, 48, 52–3, 393 data-driven journalism 42, 71–2, 364 debriefing 82, 84, 89, 90–1, 162, 313 deception 82, 85, 88, 89–90, 91, 93, 237, 245, 251 deductive reasoning see inductive versus deductive reasoning deliberate sampling 188, 203, 205–6, 207, 209, 309 demand characteristics 137, 238, 243, 251 degrees of freedom 62, 38–9, 394, 400–1, 407, 409, 410–1 dependent variable see variable depth interviews see qualitative/depth interviews description 42, 53, 54, 56–7, 58, 196 descriptive statistics see statistics dimension 102, 103, 104, 105–6, 107–8, 128, 145, 177, 178, 344, 350 direct observation see observation discriminant validity see validity discussion section 34, 42, 60, 62, 63, 64 dispersion 360, 365–6, 391; range 62, 109–10, 113–4, 143, 178–9, 198–9, 202, 294, 360, 365, 366, 369, 456 372–3, 375, 393, 443, 447; standard deviation 62, 113, 203, 360, 365, 368–9, 370–1, 374, 375, 391, 410, 443; variance 110, 199–200, 360, 365, 366–8, 370–1, 404, 412, 414, 428, 434, 441, 444; see also variance explained double-barreled question 154, 167, 174, 177 double-blind design 224, 237–8, 249 ecological validity see validity effect size 62, 194, 251–2, 401, 403–4, 404–5, 406, 434, 438, 444–5; see also variance explained empirical 3–4, 5, 25, 27, 41, 47, 55, 102, 106, 159 estimation 197–8, 200, 264, 360 ethnography 307–8, 312–3; see also field research exact replication see replication exemplars 338, 339, 340–1, 342, 350, 352, 353 exhaustive 109, 111, 154, 167, 170, 172–3 exiting 308, 313–4, 316 experiments 23–5, 49, 71, 88, 139, 189, 195, 204, 223–5, 226, 233–4, 236–7, 239–42, 249, 250–1, 261, 311, 406, 409, 411 expertise see authority explanation 42, 54, 55, 57 exploration 42, 53, 54, 55–6, 307, 309, 337–8 external validity see validity face-to-face survey 161, 212, 262, 266, 268; see also interview-style survey factor 56, 224, 233, 234–6, 240; analysis 102, 106–7, 134, 141 falsification 14, 28–9, 92 field experiment 224, 232–3, 311; notes 308, 311, 314–5, 316, 341, 343, 350, 351; research 307–8, 311–2, 314–7, 321 figures 42, 60, 65, 66, 69 filter and contingency questions 153, 161, 164, 165, 167–8, 262, 272, 273 floor effect see ceiling and floor effect focus group 7, 49, 191, 204, 249, 271, 308, 312, 317–9, 321, 337, 340, 343, 350–3; see also moderator frequency distribution 360, 365, 369–71, 372, 374, 376, 379; see also histogram fully structured interview see interview structure funnel 308, 325–6 future directions 42, 59, 60, 63–4; see also discussion section generalizability see validity grounded theory 337, 341–3, 344, 350 group equivalence see random assignment harm 43, 81, 82–3, 85, 86, 88–91, 93 Hawthorne effect 224, 243–4; see also threats to validity histogram 360, 369, 372–3 historical analysis 297, 338, 344–6 I n d e x history 224, 244; see also threats to validity honesty 81, 82 92–3, 94 hypotheses 14, 17, 19–20, 26–7, 29–30, 32, 33–5, 60, 61–2, 92, 108, 110, 141–2, 194, 236–7, 387–9, 391–7, 397–9, 400–1, 410; see also null hypothesis incentive 61, 84, 85, 87, 190, 191–2, 273, 275, 279, 318 independent variable see variable index 125, 128, 135, 141–2, 143, 154, 178, 437 indicator 23, 62, 102–7, 108, 114, 129, 133, 135, 137, 142, 145, 177–8, 298 indirect observation see observation inductive versus deductive reasoning 14, 26–7, 28, 342, 343 inference 205, 209, 387, 391–2 inferential statistics see statistics informed consent 81, 83–4, 88, 90–1, 93, 160, 162, 314 institutional review 81, 83, 85, 88, 91–2, 191 interaction 22, 66, 427, 431–2, 434–5, 437–8; see also moderation intercept see regression intercoder reliability 124, 288, 290, 293–4, 296, 298–9; statistics 288, 294–5, 296 internal validity see validity internet panels 262–3, 274–5 internet survey 161, 262, 272–3, 273–4, 275; see also self-administered survey interparticipant bias see bias intersubjective 3, 4–5, 41–2, 58–9, 69, 101, 103–4, 297–8 interval see levels of measurement interview structure 308, 326, 328–9; fully structured 327; semistructured 327–8; unstructured 328 interview-style survey 262, 266, 268, 272 intuition 42–4, 45, 46, 53, 155 item-total reliability see reliability journal article xxi, 9, 42, 61–2, 64, 66, 67–8, 72, 241 Krippendorff’s Alpha see intercoder reliability, statistics knowledge 4–5, 26–7, 29–30, 41–2, 43, 45–7, 52, 53, 59, 64, 93, 392; see also authority; empirical; intuition; tenacity kurtosis 360, 371, 375, 376, 378 latent content 288, 297, 298, 299, 349; see also manifest content leading question 154, 167, 174–5, 176 leptokurtic see kurtosis levels of measurement 102, 109–10, 110–1, 294, 361–2, 364, 405; interval 109–10, 113–4, 178, 361, 372, 410, 413, 434, 438; nominal 109, 111–2, 170, 364, 406, 409, 411, 434, 438; ordinal 110, 112–3, 296, 362–3, 366, 373; ratio 110, 114–5, 361, 366, 373, 401, 406, 409, 415, 434, 436, 438, 445 Likert-type item 114, 128, 154, 160, 178 linearity 414, 415–7, 438, 442 literature review 42, 59, 60, 61, 63–4, 93; article 42, 68, 69–70 logistic regression 427–8, 445–6 longitudinal design 52, 261–2, 265, 276–7; see also cohort; panel; trend mail survey 190, 262, 272–3, 275–6; see also self-administered survey manifest content 288, 297–8, 299; see also latent content MANOVA 427, 434, 436–8; see also ANOVA margin of error 187, 200, 163–4 matched (or paired) assignment 224, 226–7 maturation 224, 245–6; see also threats to validity mean see central tendency measurement error 123–5, 127, 153; see also random error; systematic, error median see central tendency mediation 14, 17–8, 20–1, 427 mesokurtic see kurtosis meta-analysis 42, 68, 70–1, 404 method section 42, 59, 60, 61–2 Milgram, S 44, 88–9, 91 mode see central tendency moderation 14, 17–8, 22–3, 248, 427; see also interaction moderator 308, 317–9, 320–1, 353; see also focus group mortality 137, 224, 246–7, 274, 279; see also threats to validity multivariate analysis 17, 387, 415, 427–8, 431; see also ANOVA; ANCOVA, MANOVA, regression mutually exclusive 109, 111, 154, 167, 170, 173–4 natural settings 139, 225, 241, 307, 312, 315–6 negative relationship see relationship network sampling see snowball sampling nominal see levels of measurement nonprobability sampling 188–9, 197, 203–4, 208 nonreactive measures see reactive measures normal distribution 360, 365, 368, 371, 373–4, 375, 378, 394, 440 null hypothesis 34, 388, 391, 392–3, 394–7, 397–8, 400–3, 406–7, 409–10, 412, 432–3, 434 observation 3, 23, 28, 62, 128, 136, 153–4, 155, 157, 159, 237, 266, 308, 311, 313–6, 337, 341; direct and indirect 155–6, 160; see also participant observation observer effect see Hawthorne effect observer-as-participant see participant observation odds ratio 428, 446–7; see also logistic regression 457 I n d e x one-tailed versus two-tailed test 388, 401–2, 403, 410 open-ended 107, 154, 166, 170, 171–2, 173, 237, 266, 287, 290, 296; see also closed-ended operational definition 102, 104, 105, 107–8, 123, 142, 289, 294, 299; see also conceptual definition order effects 153, 162, 164, 169 ordinal see levels of measurement others’ reports 153, 157 outliers 360–2, 376, 376–7, 378 p-value 62, 65, 388–9, 391–2, 393–5, 397, 398, 400–2, 406–7, 409, 410–1, 432, 434, 437; see also statistical significance paired assignment see matched (or paired) assignment paired t-test see t-test panel 24, 262, 271, 276, 278–9, 346; see also longitudinal design parallel-forms reliability see reliability parameter 189, 196, 198–203, 374, 387, 392 parsimony 14, 26, 30 partial replication see replication participant observation 84, 307, 314, 321–2, 323; complete observer 323; complete participant 291–2, 301; observer-as-participant 324; participant-as-observer 324 participant-as-observer see participant observation peer review 59, 66, 67, 68–9, 71–2 percent agreement see intercoder reliability, statistics periodicity 214–5 plagiarism 82, 93–4 planned comparison see post hoc test platykurtic see kurtosis polls 58, 136, 175, 187, 190, 198, 200, 208, 261, 262–3, 263–4, 265, 269, 388; tracking poll 263, 269 population 58, 71, 126, 132, 137, 139, 187–90, 193–4, 195–6, 196, 197–206, 208, 209, 211, 213, 226, 242, 246–7, 261–3, 265, 268, 287, 359–60, 373–4, 387–9, 391–4, 396, 400–1, 403, 406, 411, 413, 432, 433, 437, 439, 443 positive relationship see relationship post hoc test 194, 411–2, 434–5, 437 post-test see pre-test pre-test 169, 173, 175, 223, 239–40, 244–5, 250–1, 252 prediction 18–19; see also relationship predictive validity see validity privacy 81, 86, 87–8, 91, 268, 273, 352 probability sampling 188, 189, 198, 203, 208–9, 212–3, 213–4, 263 probing 161, 266, 268–9, 321, 326–7 publication bias see bias purposive sampling see deliberate sampling qualitative/depth interviews 87, 107, 288, 308, 311, 316–8, 324–5, 326, 337, 350–1 qualitative versus qualitative 58 458 quasi-experimental design 224, 226–7, 229–30, 233 questionnaire 64, 86, 135, 153, 160–4, 164–7, 168–70, 172, 177–8, 262, 271–6, 315 quota sampling 188–9, 206–7, 213, 274; see also nonprobability sampling R2 428, 438, 442, 444–5 random 162, 167, 169, 188, 190, 194–5, 201, 209–10, 212–5, 227, 232, 250; assignment 24, 25, 224, 225–6, 227, 229–30, 232–3, 240–2, 245, 248, 250, 252, 261, 435–6; error 52, 71, 110, 124, 125–6, 129–31, 137, 142–4, 168, 189, 190, 202, 209, 212, 228, 239, 393; number table 194, 210; sample see simple random sampling range see dispersion rapport 91, 160–2, 167, 207, 262, 266, 268, 271, 272, 275, 308, 313, 325, 328 ratio see levels of measurement reactive measures 154–5, 156, 159–60 reference list 42, 61, 64 reflexivity 307, 308–9, 313 regression 416, 427–9, 436, 438–40, 440–2, 443–5; toward the mean 224, 247–8 (see also threats to validity); see also coefficient relationship 13–7, 18, 20–6, 28–9, 33–4, 41–2, 53, 55– 8, 70, 106–7, 112, 125–6, 131–4, 137, 139, 189, 192, 193–4, 203–4, 226, 236, 241–3, 261, 263, 276, 278, 297, 361, 326, 341–3, 350, 360, 365, 369, 387–9, 391–6, 400–9, 413–6, 427–32, 433, 436–40, 442–7; positive versus negative 19–20, 33, 143, 341, 402 reliability 62, 123–6, 127–8, 130, 132, 135, 142–4, 266, 288, 327–8; item-total 128–9, 142–4; parallel-forms 129–30, 137; split-halves 130–1; test-retest 131 replication 41, 46–7, 48, 49, 52, 104, 224, 241–2 representative sample 52, 188, 189–90, 192, 197, 203–4, 206, 211, 212, 274–5 research question 14, 17, 33–4, 61–2, 103, 317, 326, 410, 412 researcher bias see bias response rate 188–9, 190–1, 197, 202, 206, 266, 271, 273–6, 278 response set 162, 166, 168–9, 262 results section 33, 42, 59, 60, 62, 69 reversed items 168 rhetorical criticism 346–8, 349–50 sample 47, 52, 60–1, 63, 91, 125, 132, 139–40, 187–9, 190–1, 193–4, 195–6, 196–215, 241, 248, 261–5, 269–79, 287, 289, 359–64, 366, 374, 387–9, 391–6, 400, 410; size 192, 193–4, 403, 410 sampling: distribution 187, 198, 200–2, 203, 211, 374, 388, 391, 393–4, 400–1; error 188–9, 192, 196–7, 200–1, 202, 209, 212, 403; frame 188, 190–1, 194–5, 209–12, 214–5, 269, 274; interval 214–5; see also systematic, sampling I n d e x saturation 308, 313, 316–7, 342 scope 14, 26, 30, 32, 224 Scott’s pi see intercoder reliability, statistics secondary analysis 41, 51–2, 71, 86, 298 selection bias see bias self-administered survey 153, 161, 167, 262, 266, 272–3, 274–5 self-correcting see cyclical and self-correcting self-report 153, 155, 156–7, 175 semantic differential item 154, 178–9 semistructured interview see interview structure sensitization 224, 228–9, 239–40, 242, 250–1; see also threats to validity significance: practical 389, 403; statistical 52, 71, 92, 391, 394, 395, 397, 397–9, 400–1, 432–3 simple random sampling 188–90, 192, 194, 195–6, 197–8, 200–6, 209, 211–5, 262, 270–1, 277, 289; see also probability sampling skew 360–2, 371–2, 374, 378, 379 snowball sampling 188, 207; see also nonprobability sampling social artifacts 154, 160, 287, 288, 289, 296, 297–8, 348 social desirability 143, 154, 157, 174, 175–6, 266 Solomon four-group design 224, 239, 240, 252 split-halves reliability see reliability spuriousness 23–5, 427, 429–30, 436; see also causality standard deviation see dispersion standard error 198, 200, 202–3, 369, 374, 439, 442 standardized coefficient see coefficient statistical control 229, 387, 427, 428–9 statistical power 193–4, 224, 193, 400–1 statistics 17, 58, 65, 70, 113, 209, 211, 341, 359–60, 428; descriptive 62, 189, 196, 360, 361–72, 376, 378, 391–2; inferential 197–8, 200, 203, 226, 387–9, 397, 401, 406 stimulus 71, 139, 223–7, 229, 230, 232, 233–4, 237, 239–41, 243 stratified sampling 188, 203, 213; see also probability sampling suppression 431 survey 261 systematic 3, 4, 41, 158, 287, 340; error 124–5, 126, 131–2, 137, 139, 143, 190, 197; sampling 188, 214 t-tests 369, 387, 389, 391, 393, 397, 401–3, 406, 409–11, 412, 438; see also bivariate analysis tables 42, 60, 65–6, 69, 203 telephone survey 131, 262, 268–9; see also interviewstyle survey tenacity 42, 46, 53 test-retest reliability see reliability testing effect 251; see also threats to validity textual analysis 338, 346, 348–50 text see social artifacts themes 317, 338, 340, 342, 345, 349, 350–1 theory 5, 8, 14, 18, 20, 22–3, 25–6, 27–34, 54–5, 62, 70, 104, 106, 350, 388, 392, 412, 437–8 third variable see variable threats to validity 242 time order see causality tracking poll see polls transcript 325, 337–8, 342–3, 348, 351–3 treatment group 223–4, 233, 237–41, 406 trend 52, 262, 276, 277–8; see also longitudinal design triangulation 41, 49, 52, 104, 139, 261, 312, 322 Tuskegee 82–3 two-tailed test see one-tailed versus two-tailed test type I error 389, 392, 395, 396–7, 397–99, 400, 411–2, 435, 437; see also Bonferroni correction; statistical significance type II error 193, 389, 396, 400–1, 432; see also statistical significance unit of analysis 13, 15, 16, 109, 153, 157, 211, 289–91, 307, 343 unit of observation 15; see also unit of analysis unitizing 288–9, 343 unstandardized coefficient see coefficient unstructured interview see interview structure validity 8, 43, 84, 108–9, 123–6, 127, 131–2, 153, 161, 168–9, 174–5, 177, 269, 272, 291, 327–8; criterion (concurrent and predictive) 130, 136–7; construct (convergent and discriminant) 132–4; content 135–6; ecological validity 224, 242; external validity 139–40, 195, 204, 224, 229, 232, 234, 241–6, 251–2, 261; internal validity 137–8, 195, 204, 224, 241–6, 251–2 variable 13–4, 18–9, 20, 22–3, 25–6, 29–30, 33–4, 52–3, 53–4, 55–8, 62, 70, 102–3, 108–10, 111, 112–3, 123–8, 131–7, 141–2, 153–4, 178, 204, 226, 227, 261, 287, 289–90, 294, 316, 359–69, 372–3, 387, 392, 395, 403, 405–7, 413–4, 416; control 21, 176, 428; dependent 16, 18, 20, 22–5, 94, 223, 227, 229, 233–4, 239, 244, 276, 278, 397, 407, 409, 411, 414–5, 427, 431, 434, 436–40, 443–6; independent 15–6, 18, 20, 22–5, 94, 223, 232–5, 244, 276, 278, 404, 407, 409, 411,414–5, 427, 431, 434, 436–40, 443–6; third 17, 20, 22–5, 71, 225, 228, 230, 248, 404, 427, 429–31 variance see dispersion variance explained 389, 404–5, 412, 428–9, 442, 444, 445 voluntary participation 81–2, 84–5, 275, 314 volunteer sampling 188, 203, 205, 206, 207–8 within-subjects design 224, 227, 228–9, 239 zero-order see bivariate analysis 459 Taylor & Francis eBooks www.taylorfrancis.com A single destination for eBooks from Taylor & Francis with increased functionality and an improved user experience to meet the needs of our customers 90,000+ eBooks of award-winning academic content in Humanities, Social Science, Science, Technology, Engineering, and Medical written by a global network of editors and authors TAYLOR & FRANCIS EBOOKS OFFERS: A streamlined experience for our library customers A single point of discovery for all of our eBook content Improved search and discovery of content at both book and chapter level REQUEST A FREE TRIAL support@taylorfrancis.com ... e p t s o f r e s e a r c h Consider an instance in which an advertising agency knows that a particular advertising approach shares a causal relationship with sales of a particular type of product... Unit replication ® Unit Systematic Research is about more than observing and measuring It takes a systematic approach to gathering and analyzing data, which means that researchers follow a set... 1975– author | Schmierbach, Mike, 1976– author Title: Applied communication research methods : getting started as a researcher / Michael P Boyle, Mike Schmierbach Description: Second edition |