Appendix A



Appendix ASummary of Protection Motivation ResearchTable A1.1 Summary of Protection Motivation ResearchReferenceNSample frameThreat TargetResultsAnderson and Agarwal, 2010 [1]594Undergraduate students and internet subscribers in a rural areaHome users“The greater the sense of ownership an individual feels for his/her computer, the higher are his/her intentions to protect it.” P. 628Boss et al., 2015 [2]S1: 104S2: 327S1: MBA students in an introductory information systems courseS2: Psychology students in a large universityStudentsAll level of messages: only response efficacy and response cost showed significant impact on intent.High level threat messages: all paths significant to intentClaar and Johnson, 2012 [3]184Snowball sample starting with undergraduate students at a Western US universityStudents and undefinedSeverity not supported.Crossler and Belanger, 2014 [4]81Non-information systems graduate studentsStudentsResponse costs showed no significance on the dependent variable (unified security practices). Crossler et al., 2014 [5]444Undergraduate and graduate students, white collar workersMixedThreat susceptibility and response cost not supported. Self-efficacy and response efficacy showed significant impact on compliance.Gurung et al., 2009 [6]232Business undergraduate students at three different universitiesStudentsNeither vulnerability nor response costs showed significant impact on use of anti-spyware tools.Herath and Rao, 2009 [7]312High-level information systems managers in approximately 690 organizationsEmployeesNeither perceived vulnerability (probability) nor severity were significant.Herath et al., 2014 [8]134Junior-level undergraduate students at a large public university in the north-east USStudentsThreat appraisal and internal coping mechanism appraisal both supported.Ifinedo, 2012 [9]124Non-information systems managers in Canadian organizations from InfoCANADA and information system professionalsEmployeesPerceived security and response cost were not predictors of information systems security policy compliance.Johnston and Warkentin, 2010 [10]215Faculty, staff, and students from multiple units at one large universityEmployees and studentsThreat susceptibility did not have significant effect on intent.Johnston et al., 2015 [11]559Government employees from a city government in FinlandEmployeesThereat susceptibility was not a significant determinant in intentions to comply with recommended protective strategies.LaRose et al., 2008 [12]206College students from an introductory mass communication class.EmployeesPersonal responsibility for Internet safety and self-efficacy demonstrated significant impact on safety intentions. Lee et al., 2008 [13]273Undergraduate students at a large midwestern US universityStudentsPerceived severity did not demonstrate significant impact on intention.Lee, 2011 [14]218Faculty members working at US universitiesFacultyResponse cost and self-efficacy not significant for actual behavior.Lee and Larsen, 2009 [15]239Small and medium-sized business executives in the USExecutivesResponse efficacy, self-efficacy, and perceived vulnerability lacked significant impact on intention to adopt anti-malware in multi-group analysis.Marett et al., 2011 [16]522Undergraduate management information system students and responses from “snowball” sample StudentsExtrinsic rewards had unexpected positive influence on the threat appraisal model and were not significant in the full PMT model.Ng et al., 2009 [17]134Employed part-time students in two computing classes at a large public university, employees of three information technology-related organizationsEmployeesPerceived severity showed no significant impact on computer security behavior.Posey et al., 2015 [18]380“Organizational insiders” in the USA recruited using an external panel providerEmployeesFear did not play a significant role in motivating users.Siponen et al., 2010 [19]917Employees in FinlandEmployeesResponse efficacy did not have a significant impact on intent to comply.Vance et al., 2012 [20]210A large organizationEmployeesVulnerability had an insignificant impact on employees.Yoon et al., 2012 [21]202Business students from a South Korean universityStudentsResponse-efficacy and self-efficacy had a strong impact on students’ intention and behavior.Zhang and McDowell, 2009 [22]182Students from three universities in the USStudentsNeither severity nor vulnerability supported.Table data partially adapted from Sommestad et al. 2015 and Boss et al. 2015REFERENCESAnderson, C.L. and Agarwal, R. Practicing Safe Computing: A Multimethod Empirical Examination of Home Computer User Security Behavioral Intentions. MIS Quarterly, 34, 3 (2010), 613–643.Boss, S.R., Galletta, D.F., Lowry, P.B., Moody, G.D., and Polak, P. What do systems users have to fear? Using fear appeals to engender threats and fear that motivate protective security behaviors. MIS Quarterly, 39, 4 (2015), 837–864.Claar, C.L. and Johnson, J. Analyzing Home PC Security Adoption Behavior. Journal of Computer Information Systems, 52, 4 (2012), 20–29.Crossler, R.E. and Bélanger, F. Determinants of Individual Security Behaviors. In Proceedings of the 2010 International Federation of Information Processing (IFIP) 8.11/11.13 Dewald Roode Workshop on Information Systems Security Research. 2010, pp. 78–127.Crossler, R.E., Long, J.H., Loraas, T.M., and Trinkle, B.S. Understanding Compliance with Bring Your Own Device Policies Utilizing Protection Motivation Theory: Bridging the Intention-Behavior Gap. Journal of Information Systems, 28, 1 (2014), 209–226.Gurung, A., Luo, X., and Liao, Q. Consumer Motivations in Taking Action Against Spyware: An Empirical Investigation. Information Management & Computer Security, 17, 3 (2008), 276–289.Herath, T. and Rao, H.R. Protection Motivation and Deterrence: A Framework for Security Policy Compliance in Organisations. European Journal of Information Systems, 18, 2 (April 2009), 106–125.Herath, T., Chen, R., Wang, J., Banjara, K., Wilbur, J., and Rao, H.R. Security services as coping mechanisms: An investigation into user intention to adopt an email authentication service. Information Systems Journal, 24, 1 (2014), 61–84.Ifinedo, P. Understanding Information Systems Security Policy Compliance: An Integration of the Theory of Planned Behavior and the Protection Motivation Theory. Computers & Security, 31, 1 (November 2012), 83–95.Johnston, A.C. and Warkentin, M. Fear Appeals and Information Security Behaviors: An Empirical Study. MIS Quarterly, 34, 3 (2010), 549–566.Johnston, A.C., Warkentin, M., and Siponen, M. An Enhanced Fear Appeal Framework: Leveraging Threats to the Human Asset through Sanctioning Rhetoric. MIS Quarterly, 39, 1 (2015), 113–134.LaRose, R., Rifon, N., and Enbody, R. Promoting personal responsibility for internet safety. Communications of the ACM, 51, 3 (2008), 71-76.Lee, D., Larose, R., and Rifon, N. Keeping our network safe: a model of online protection behaviour. Behaviour & Information Technology, 27, 5 (September 2008), 445–454.Lee, Y. Understanding anti-plagiarism software adoption: An extended protection motivation theory perspective. Decision Support Systems, 50, 2 (2011), 361–369.Lee, Y. and Larsen, K.R. Threat or Coping Appraisal: Determinants of SMB Executives’ Decision to Adopt Anti-Malware Software. European Journal of Information Systems, 18, 2 (March 2009), 177–187.Marett, K., McNab, A.L., and Harris, R.B. Social networking websites and posting personal information: An evaluation of protection motivation theory. AIS Transactions on Human-Computer Interaction, 3, 3 (2011), 170–188.Ng, B.-Y., Kankanhalli, A., and Xu, Y. (Calvin). Studying Users’ Computer Security Behavior: A Health Belief Perspective. Decision Support Systems, 46, 4 (March 2009), 815–825.Posey, C., Roberts, T., and Lowry, P.B. The Impact of Organizational Commitment on Insiders’ Motivation to Protect Organizational Information Assets. Journal of Management Information Systems, 32, 4 (2015), 179-214.Siponen, M., Pahnila, S., and Mahmood, M.A. Compliance with information security policies: An empirical investigation. Computer, 43, 2 (2010), 64–71.Vance, A., Siponen, M., and Pahnila, S. Motivating IS security compliance: Insights from Habit and Protection Motivation Theory. Information & Management, 49, 3–4 (May 2012), 190–198.Yoon, C., Hwang, J., and Kim, R. Exploring factors that influence students’ behaviors in information security. Journal of Information Systems Education, 23, 4 (2012), 407–415.Zhang, L. and McDowell, W.C. Am I really at risk? Determinants of online users’ intentions to use strong passwords. Journal of Internet Commerce, 8, 3–4 (2009), 180–197.Appendix BEnd User Appeal ConstructionBaseline AppealPlease read the following message very carefully:[RELATEDNESS STATEMENT] Hackers often accomplish identity theft by figuring out online passwords. [COMPETENCE STATEMENT] A password manager is software that aids in keeping track of multiple passwords. [SEVERITY STATEMENT] [SUSCEPTIBILITY STATEMENT] [RESPONSE EFFICACY STATEMENT] [SELF-EFFICACY STATEMENT] [RESPONSE COST STATEMENT] We recommend using Dashlane [AUTONOMY STATEMENT] as your password manager.Manipulation StatementsRelatedness Statement: Your passwords are the keys to your digital life, and your online accounts are a proverbial gold mine for someone looking to steal your petence Statement: Regardless of how confident you are in your computer skills, you can learn how to create strong passwords and manage them using a password manager.Severity Statement: By obtaining a password, hackers can collect almost any type of data, such as additional user logins, bank accounts, credit card information and gain control of social media accounts.Susceptibility Statement: Companies that you may do business with, such as Sony or Apple, keep track of passwords and have suffered previous attacks. Your password could be compromised through such an attack.Response Efficacy Statement: A password manager assists in generating a unique secure password for each of your accounts, so that if an attack occurs, only one of your accounts is affected.Self-efficacy Statement: A password manager is very easy to install and useResponse Cost Statement: …does not affect the usability of your computer, and is often provided for free.Autonomy Statement: …1Password, KeePass, or LastPass. Each of these is an adequate solution, so feel free to choose the software you like the bestExample Self-Determined AppealYour passwords are the keys to your digital life, and your online accounts are a proverbial gold mine for someone looking to steal your identity. Hackers often accomplish identity theft by figuring out online passwords. Regardless of how confident you are in your computer skills, you can learn how to create strong passwords and manage them using a password manager. A password manager is software that aids in keeping track of multiple passwords. We recommend using Dashlane, 1Password, KeePass, or LastPass. Each of these is an adequate solution, so feel free to choose the software you like the best as your password manager.Example Fear AppealHackers often accomplish identity theft by figuring out online passwords. A password manager is software that aids in keeping track of multiple passwords. By obtaining a password, hackers can collect almost any type of data, such as additional user logins, bank accounts, credit card information and gain control of social media accounts. Companies that you may do business with, such as Sony or Apple, keep track of passwords and have suffered previous attacks. Your password could be compromised through such an attack. A password manager assists in generating a unique secure password for each of your accounts, so that if an attack occurs, only one of your accounts is affected. A password manager is very easy to install and use, does not affect the usability of your computer, and is often provided for free. We recommend using Dashlane as your password manager.Appendix CInstrument ItemsBehavioral Intention to Install Password Manager SoftwareNow that you have read the message above, please indicate the likelihood that you will install the password manager software described above:Likelihood to install slider scale, 0–10 (0 = Extremely unlikely; 10 = Extremely likely)Response Performance Motivation (I would choose to install password manager software…)…because I think that this activity is interesting.…because I think that this activity is pleasant.…because I think that this activity is fun.…because I feel good when doing this activity.…because I am doing it for my own good.…because I think that this activity is good for me.…because I decided that this activity is beneficial.…because I believe that this activity is important to me.…because I am supposed to do it.…because it is something that I have to do.…because I don’t have any choice.…because I feel that I have to do it.…but I am not sure if it is worth it.…but I don’t see what the activity brings me.…but I am not sure it is a good thing to pursue it.…but personally I don’t see any good reasons to do this activity.Threat SeverityIf my online passwords were discovered by hackers, it would be severe.If my online passwords were discovered by hackers, it would be serious.If my online passwords were discovered by hackers, it would be significant.Threat SusceptibilityMy online passwords are at risk for becoming compromised.It is likely that my online passwords will be breached.It is possible that my online passwords will be compromised.Response EfficacyPassword manager software works for protection.Password manager software is effective for protection.When using password manager software, online accounts are more likely to be protected.Self-EfficacyPassword manager software is easy to use.Password manager software is convenient to use.I am able to use a password manager without much effort.Response CostUsing password manager software is time consuming for me. (formative)Using password manager software is burdensome for me. (formative)Using password manager software is financially costly for me. (formative)Installing password manager software would require too much from me. (reflective)Installing password manager software is not worth it. (reflective)AutonomyThe software described is what I would choose to install on my computer.I feel that the software I’m told to install fits perfectly with what I prefer to use on my computer.I feel that the software described is an expression of my own software preferences.I feel that I have the opportunity to make choices with respect to what I am told to install in the petenceI feel I have a better understanding of password manager software.I feel that I effectively learned about password manager software.I feel that I did a good job learning about password manager software.I feel that I can manage the requirements of learning more about password manager software.RelatednessI feel a strong connection with my digital information.If the information contained in my online accounts is affected, then so am I.The thought of information contained in my online accounts being tampered with makes me anxious.Protecting the information contained in my online accounts is a way to protect puting ExperienceHow many total years of general experience do you have working with a computer in any form (e.g., surfing the internet, spreadsheets, gaming, word processing)? (Text entry)Appendix DValidity Testing and Post Hoc AnalysesPanel and Pilot TestingAn expert panel composed of professors and graduate students experienced in both behavioral information security research and experimental research designs examined the structure of the appeal and the measurement items. After adjusting the instrument based on the panel’s recommendations, we pilot-tested the instrument using a small sample from Amazon Mechanical Turk (n = 104). The pilot test revealed no initial validity problems, and we proceeded to our full data collection.Data Validation Practices for Mechanical TurkAmazon Mechanical Turk has been previously validated as a source for recruiting respondents for inclusion in academically rigorous research ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Steelman", "given" : "Zachary R", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hammer", "given" : "Bryan I", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Limayem", "given" : "Moez", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "MIS Quarterly", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2014" ] ] }, "page" : "355-378", "title" : "Data Collection in the Digital Age: Innovative Alterantives to Student Samples", "type" : "article-journal", "volume" : "38" }, "uris" : [ "", "" ] } ], "mendeley" : { "formattedCitation" : "[13]", "plainTextFormattedCitation" : "[13]", "previouslyFormattedCitation" : "[68]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[13]. To further ensure valid responses from our sample, we followed a series of recommended guidelines for demonstrating valid responses ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Crump", "given" : "Matthew J C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "V", "family" : "McDonnell", "given" : "John", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Gureckis", "given" : "Todd M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "PLoS ONE", "id" : "ITEM-1", "issue" : "3", "issued" : { "date-parts" : [ [ "2013" ] ] }, "page" : "e57410", "title" : "Evaluating Amazon's Mechanical Turk as a Tool for Experimental Behavioral Research", "type" : "article-journal", "volume" : "8" }, "uris" : [ "", "" ] } ], "mendeley" : { "formattedCitation" : "[3]", "plainTextFormattedCitation" : "[3]", "previouslyFormattedCitation" : "[20]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[3]. We exclusively recruited our respondents from the United States; US respondents are shown to be more reliable than respondents recruited from other countries. We embedded attention filter questions among our measurement items (i.e., “Please answer disagree for this statement”) to ensure respondents were not systematically answering scale items (e.g., answering agree for all items). Finally, we excluded any responses that were completed in an unreasonably fast amount of time (difference greater than 3 standard deviations from the mean completion time).Demographic InformationOur sample represented a fairly even split of males and females (282 males; 261 females; 2 unreported). White, non-Hispanic respondents comprised the majority of our sample (66.0%), followed by Hispanic (16.8%), Asian (8.4%), African American (5.5%), Native American (0.4%), and Pacific Islander (0.4%) respondents. The remaining respondents answered “other” (2.2%) or did not report ethnicity (0.4%). The average age of our respondents was 33.6 years (SD = 10.42), while the average amount of computing experience was 18.77 years (SD = 6.37).Instrument ValidityWe adapted previously validated multi-item scales to measure constructs modeled as reflective (because motivation to perform the recommended response is measured as an index and response cost is modeled as a formative construct, these are not included in our analysis of convergent and discriminant validity). According to Loch et al. ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "Drawing on the theoretical work of Hill et al. and Straub et al., this study examines culture-specific inducements and impediments to using the Internet in the Arab world. Research questions were 1) to what extent does the process of technology cul- turation affect the acceptance of the Internet 2) to what extent do social norms (SNs) affect the acceptance of the Internet? Of the two research methods employed, the first was a quanti- tative field study of knowledgeworkers. The instrument measured the extent to which respondents and their organizations are influ- enced by advanced technology cultures. Using partial least squares (PLS), the first of two models tested links between SNs; techno- logical culturation and Internet usage for each respondent. The second model investigated links between technological culturation and Internet utilization for the respondent\u2019s organization. Find- ings showstrong support for both models, explaining, respectively, 47% and 37% of the variance. The second method was a quali- tative analysis of respondents\u2019 free-format comments. These find- ings reinforce the quantitative findings, on the one hand, and re- veal additional cultural barriers that still need to be studied, on the other. Findings identify how culture can both inhibit and en- courage technological innovation and howArab cultures can move their economies more quickly into the digital age.", "author" : [ { "dropping-particle" : "", "family" : "Loch", "given" : "Karen D", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Straub", "given" : "Detmar W", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kamel", "given" : "Sherif", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "IEEE Transactions on Engineering Management", "id" : "ITEM-1", "issue" : "1", "issued" : { "date-parts" : [ [ "2003", "2" ] ] }, "page" : "45-63", "title" : "Diffusing the Internet in the Arab World: The Role of Social Norms and Technological Culturation", "type" : "article-journal", "volume" : "50" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[6]", "plainTextFormattedCitation" : "[6]", "previouslyFormattedCitation" : "[44]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[6], convergent validity must be established to ensure that each item measuring a particular construct is significantly correlated with its construct’s composite value. We followed the latest standards for establishing factorial validity in PLS ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "This tutorial explains in detail what factorial validity is and how to run its various aspects in PLS. The tutorial is written as a teaching aid for doctoral seminars that may cover PLS and for researchers interested in learning PLS. An annotated example with data is provided as an additional tool to assist the reader in reconstructing the detailed example.", "author" : [ { "dropping-particle" : "", "family" : "Gefen", "given" : "David", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Straub", "given" : "Detmar W", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Communications of the Association for Information Systems", "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2005" ] ] }, "page" : "91-109", "title" : "A Practical Guide to Factorial Validity Using PLS-Graph: Tutorial and Annotated Example", "type" : "article-journal", "volume" : "16" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Lowry", "given" : "Paul Benjamin", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Gaskin", "given" : "James", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "IEEE Transactions on Professional Communication", "id" : "ITEM-2", "issue" : "2", "issued" : { "date-parts" : [ [ "2014" ] ] }, "page" : "123-146", "title" : "Partial least squares (PLS) structural equation modeling (SEM) for building and testing behavioral causal theory: When to choose it and how to use it", "type" : "article-journal", "volume" : "57" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[5,7]", "plainTextFormattedCitation" : "[5,7]", "previouslyFormattedCitation" : "[30,45]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[5,7]. In examining PLS reports for cross-loadings, convergent validity was significantly established for all constructs except autonomy and competence. Items that did not load with their respective construct (AUTO4 and COMP4) were subsequently removed from the analysis. In addition to item loadings above 0.70 on intended factors, an indicator of convergent validity is an average variance extracted greater than 0.50 ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "This tutorial explains in detail what factorial validity is and how to run its various aspects in PLS. The tutorial is written as a teaching aid for doctoral seminars that may cover PLS and for researchers interested in learning PLS. An annotated example with data is provided as an additional tool to assist the reader in reconstructing the detailed example.", "author" : [ { "dropping-particle" : "", "family" : "Gefen", "given" : "David", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Straub", "given" : "Detmar W", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Communications of the Association for Information Systems", "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2005" ] ] }, "page" : "91-109", "title" : "A Practical Guide to Factorial Validity Using PLS-Graph: Tutorial and Annotated Example", "type" : "article-journal", "volume" : "16" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[5]", "plainTextFormattedCitation" : "[5]", "previouslyFormattedCitation" : "[30]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[5]. An average variance extracted of 0.50 or greater was achieved on all constructs. With offending items removed, convergent validity was established for all constructs. Factor loading scores for respondents used in the analysis of our integrated model are shown in REF _Ref476839469 \h \* MERGEFORMAT Table D1. The scores for PMT and SDT measurement items for our model comparison subsamples are shown in REF _Ref444357055 \h \* MERGEFORMAT Table D2 and REF _Ref444357063 \h \* MERGEFORMAT Table D3, respectively.The reliability of the scales was evaluated using composite reliability scores. Values greater than or equal to 0.70 for these measures are recognized as acceptable ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "Despite the fact that validating the measures of constructs is critical to building cumulative knowledge in MIS and the behavioral sciences, the process of scale development and validation continues to be a challenging activity. Undoubtedly, part of the problem is that many of the scale development procedures advocated in the literature are limited by the fact that they (1) fail to adequately discuss how to develop appropriate conceptual definitions of the focal construct, (2) often fail to properly specify the measurement model that relates the latent construct to its indicators, and (3) underutilize techniques that provide evidence that the set of items used to represent the focal construct actually measures what it purports to measure. Therefore, the purpose of the present paper is to integrate new and existing techniques into a comprehensive set of recommendations that can be used to give researchers in MIS and the behavioral sciences a framework for developing valid measures. First, we briefly elaborate upon some of the limitations of current scale development practices. Following this, we discuss each of the steps in the scale development process while paying particular attention to the differences that are required when one is attempting to develop scales for constructs with formative indicators as opposed to constructs with reflective indicators. Finally, we discuss several things that should be done after the initial development of a scale to examine its generalizability and to enhance its usefulness.", "author" : [ { "dropping-particle" : "", "family" : "Mackenzie", "given" : "Scott B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Philip M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Nathan P", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "MIS Quarterly", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2011" ] ] }, "page" : "293-334", "title" : "Construct Measurement and Validation Procedures in MIS and Behavioral Research: Integrating New and Existing Techniques", "type" : "article-journal", "volume" : "35" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "abstract" : "A critical element in the evolution of a fundamental body of knowledge in marketing, as well as for improved marketing practice, is the development of better measures of the variables with which marketers work. In this article an appraach is outlined by which this goal can be achieved and portions of the approach are illustrated in terms of a job satisfaction measure.", "author" : [ { "dropping-particle" : "", "family" : "Churchill", "given" : "Gilbert A", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Marketing Research", "id" : "ITEM-2", "issue" : "February", "issued" : { "date-parts" : [ [ "1979" ] ] }, "page" : "64-73", "title" : "A Paradigm for Developing Better Measures of Marketing Constructs", "type" : "article-journal", "volume" : "16" }, "uris" : [ "", "" ] }, { "id" : "ITEM-3", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Peter", "given" : "J Paul", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Marketing Research", "id" : "ITEM-3", "issue" : "2", "issued" : { "date-parts" : [ [ "1981" ] ] }, "page" : "133-145", "title" : "Construct Validity: A Review of Basic Issues and Marketing Practices", "type" : "article-journal", "volume" : "18" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[2,8,9]", "plainTextFormattedCitation" : "[2,8,9]", "previouslyFormattedCitation" : "[14,46,56]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[2,8,9]. All constructs achieved scores of at least 0.70 for composite reliability, as demonstrated in REF _Ref476839469 \h \* MERGEFORMAT Table D1, REF _Ref444357055 \h \* MERGEFORMAT Table D2, and REF _Ref444357063 \h \* MERGEFORMAT Table D3.Table D SEQ Table \* ARABIC 1: Cross-Loadings, Reliability, and AVE for Latent Variables - Integrated Model AUTOCOMPREFRELSEFTSEVTSUSComposite ReliabilityAVEAUTO1.931.323.588.212.340.214.196.950.865AUTO2.945.366.599.193.395.179.190AUTO3.913.303.519.166.334.151.153COMP1.386.923.326.134.232.035.026.947.857COMP2.312.934.225.096.247.021.001COMP3.279.921.182.067.233.001?.025REF1.518.264.876.166.258.158.187.918.788REF2.594.274.911.194.307.157.171REF3.516.177.875.165.251.126.169REL1.213.081.153.731.177.460.171.875.636REL2.103.078.139.834.194.531.310REL3.213.084.152.807.163.541.381REL4.127.107.189.814.172.511.273SEF1.275.243.247.182.854.164.093.895.739SEF2.386.218.316.208.862.231.097SEF3.313.203.222.175.862.165.062SEV1.184.008.145.583.192.919.342.945.850SEV2.190.029.157.621.213.930.320SEV3.168.025.157.572.206.918.320SUS1.235?.001.261.266.065.277.837.850.653SUS2.188.042.107.254.098.257.772SUS3.058?.030.115.354.076.323.815Table D SEQ Table \* ARABIC 2: Cross-Loadings, Reliability, and AVE for Latent PMT VariablesREFSEFSEVSUSComposite ReliabilityAVEREF1.898.270.214.133.929.814REF2.915.359.225.143REF3.893.299.193.109SEF1.289.860.201.042.882.713SEF2.289.839.231.109SEF3.294.835.111.013SEV1.206.170.908.339.949.862SEV2.217.216.943.294SEV3.226.211.934.361SUS1.157.076.329.930.814.600SUS2.062.043.221.738SUS3.079.085.291.625REF = Response Efficacy; SEF = Self-efficacy; SEV = Threat Severity; SUS = Threat SusceptibilityTable D SEQ Table \* ARABIC 3: Cross-Loadings, Reliability, and AVE for Latent SDT VariablesAUTOCOMPRELComposite ReliabilityAVEAUTO1.933.300.259.946.853AUTO2.934.284.244AUTO3.904.244.211COMP1.327.922.124.946.854COMP2.268.940.085COMP3.226.909.077REL1.216.061.765.888.665REL2.156.080.863REL3.274.101.775REL4.203.100.856AUTO = Perceived Autonomy; COMP = Perceived Competence; REL = Perceived RelatednessWe procedurally addressed the concerns of common method bias ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "ISSN" : "0021-9010", "abstract" : "Interest in the problem of method biases has a long history in the behavioral sciences. Despite this, a comprehensive summary of the potential sources of method biases and how to control for them does not exist. Therefore, the purpose of this article is to examine the extent to which method biases influence behavioral research results, identify potential sources of method biases, discuss the cognitive processes through which method biases influence responses to measures, evaluate the many different procedural and statistical techniques that can be used to control method biases, and provide recommendations for how to select appropriate procedural and statistical remedies for different types of research settings.", "author" : [ { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Philip M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "MacKenzie", "given" : "Scott B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Lee", "given" : "Jeong-Yeon", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Nathan P", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Applied Psychology", "id" : "ITEM-1", "issue" : "5", "issued" : { "date-parts" : [ [ "2003", "10" ] ] }, "page" : "879-903", "title" : "Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies", "type" : "article-journal", "volume" : "88" }, "uris" : [ "", "" ] }, { "id" : "ITEM-2", "itemData" : { "DOI" : "10.1146/annurev-psych-120710-100452", "ISSN" : "1545-2085", "PMID" : "21838546", "abstract" : "Despite the concern that has been expressed about potential method biases, and the pervasiveness of research settings with the potential to produce them, there is disagreement about whether they really are a problem for researchers in the behavioral sciences. Therefore, the purpose of this review is to explore the current state of knowledge about method biases. First, we explore the meaning of the terms \"method\" and \"method bias\" and then we examine whether method biases influence all measures equally. Next, we review the evidence of the effects that method biases have on individual measures and on the covariation between different constructs. Following this, we evaluate the procedural and statistical remedies that have been used to control method biases and provide recommendations for minimizing method bias.", "author" : [ { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Philip M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "MacKenzie", "given" : "Scott B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Nathan P", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Annual Review of Psychology", "id" : "ITEM-2", "issued" : { "date-parts" : [ [ "2012", "1" ] ] }, "page" : "539-569", "title" : "Sources of method bias in social science research and recommendations on how to control it", "type" : "article-journal", "volume" : "63" }, "uris" : [ "", "" ] } ], "mendeley" : { "formattedCitation" : "[10,11]", "plainTextFormattedCitation" : "[10,11]", "previouslyFormattedCitation" : "[57,58]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[10,11] by ensuring respondents of their anonymity and engaging an expert panel to ensure items were clear, succinct, and precise ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "Despite the fact that validating the measures of constructs is critical to building cumulative knowledge in MIS and the behavioral sciences, the process of scale development and validation continues to be a challenging activity. Undoubtedly, part of the problem is that many of the scale development procedures advocated in the literature are limited by the fact that they (1) fail to adequately discuss how to develop appropriate conceptual definitions of the focal construct, (2) often fail to properly specify the measurement model that relates the latent construct to its indicators, and (3) underutilize techniques that provide evidence that the set of items used to represent the focal construct actually measures what it purports to measure. Therefore, the purpose of the present paper is to integrate new and existing techniques into a comprehensive set of recommendations that can be used to give researchers in MIS and the behavioral sciences a framework for developing valid measures. First, we briefly elaborate upon some of the limitations of current scale development practices. Following this, we discuss each of the steps in the scale development process while paying particular attention to the differences that are required when one is attempting to develop scales for constructs with formative indicators as opposed to constructs with reflective indicators. Finally, we discuss several things that should be done after the initial development of a scale to examine its generalizability and to enhance its usefulness.", "author" : [ { "dropping-particle" : "", "family" : "Mackenzie", "given" : "Scott B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Philip M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Podsakoff", "given" : "Nathan P", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "MIS Quarterly", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2011" ] ] }, "page" : "293-334", "title" : "Construct Measurement and Validation Procedures in MIS and Behavioral Research: Integrating New and Existing Techniques", "type" : "article-journal", "volume" : "35" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[8]", "plainTextFormattedCitation" : "[8]", "previouslyFormattedCitation" : "[46]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[8]. We analytically assessed our data for the presence of common method variance (CMV) using an unmeasured latent construct using an analysis of moment structures (AMOS). A χ2 difference of 3.84 or more between the measurement model and the model containing the unmeasured latent method construct indicates common method variance has a significant impact on the model. The difference between our models was 0.744 (measurement model χ2 = 352.957; ULMC model χ2 = 352.213), indicating that CMV did not significantly affect our dataset. When using reflective measures, discriminant validity should also be demonstrated. Cross-loadings exceeding 0.40 were not indicated for any items. We further establish construct validity by analyzing the average variance extracted value, compared with the shared variance between constructs. All constructs demonstrated discriminant validity. REF _Ref476839703 \h \* MERGEFORMAT Table D4 shows the correlations of variables for our integrated model subset. REF _Ref444357229 \h \* MERGEFORMAT Table D5 and REF _Ref444357212 \h \* MERGEFORMAT Table D6 illustrate inter-construct correlations for latent constructs related to PMT and SDT model comparison subsets, respectively.Table D SEQ Table \* ARABIC 4: Mean, Standard Deviation, and Correlations for Latent Constructs – Integrated Model?MeanSDAUTOBICOMPCOSMOTREFRELSEFSEVSUSAUTO2.770.96(.930)BI3.652.76.662(----)COMP3.300.96.356.365(.926)COS2.790.84?.351?.338?.225(----)MOT2.840.59.617.643.296-.404(----)REF3.440.85.613.532.270-.281.562(.887)REL3.960.67.206.229.110-.144.235.198(.798)SEF3.300.73.384.337.256-.557.345.308.220(.860)SEV3.910.89.196.219.022-.122.217.166.642.221(.922)SUS3.220.78.194.171.003.055.175.198.363.098.355(.808)Table D SEQ Table \* ARABIC 5: Mean, Standard Deviation, and Correlations for Reflective Latent Constructs – PMT-Only ModelMeanStandard DeviationREFSEFSEVSUSREF3.4290.887(.902)SEF3.2700.700.344(.844)SEV3.9500.932.234.215(.928)SUS3.3390.730.142.056.357(.775)( ) = square-root AVE; REF = Response Efficacy; SEF = Self-efficacy; SEV = Threat Severity; SUS = Threat SusceptibilityTable D SEQ Table \* ARABIC 6: Mean, Standard Deviation, and Correlations for Reflective Latent Constructs – SDT-Only ModelMeanStandard DeviationAUTOCOMPRELAUTO2.7550.918(.924)COMP3.3750.909.280(.924)REL3.9330.689.148.088(.815)( ) = square-root AVE; AUTO = Perceived Autonomy; COMP = Perceived Competence; REL = Perceived RelatednessBecause our development of a self-determined appeal featured novel and previously untested manipulation statements directly related to autonomy, competence, and relatedness, it is necessary to demonstrate the effectiveness of the statements. We compared mean scores of latent construct perceptions based on whether the treatment was received. Each appeal treatment significantly demonstrated its intended effect on respondents’ perceptions. Results of the independent samples t-tests are further shown in REF _Ref476839893 \h \* MERGEFORMAT Table D7.Table D SEQ Table \* ARABIC 7: Independent Samples t-Test Results for Appeal TreatmentsAppeal TreatmentWithout TreatmentWith Treatmentt-Statp-Value Mean SD Mean SDAutonomy2.6800.9422.8490.9812.046.020Competence3.2100.9983.3850.9152.134.016Relatedness3.9000.6864.0250.6532.175.015Post Hoc AnalysesMediation AnalysisBecause the integrated SDT-PMT model depicts several mediated relationships, we also conducted a series of mediation tests to assess the nature of the indirect effects shown in the model. Based on guidelines for mediation testing specified by Baron and Kenny ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "ISSN" : "0022-3514", "PMID" : "3806354", "abstract" : "In this article, we attempt to distinguish between the properties of moderator and mediator variables at a number of levels. First, we seek to make theorists and researchers aware of the importance of not using the terms moderator and mediator interchangeably by carefully elaborating, both conceptually and strategically, the many ways in which moderators and mediators differ. We then go beyond this largely pedagogical function and delineate the conceptual and strategic implications of making use of such distinctions with regard to a wide range of phenomena, including control and stress, attitudes, and personality traits. We also provide a specific compendium of analytic procedures appropriate for making the most effective use of the moderator and mediator distinction, both separately and in terms of a broader causal system that includes both moderators and mediators.", "author" : [ { "dropping-particle" : "", "family" : "Baron", "given" : "Reuben M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kenny", "given" : "David A", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Personality and Social Psychology", "id" : "ITEM-1", "issue" : "6", "issued" : { "date-parts" : [ [ "1986", "12" ] ] }, "page" : "1173-1182", "title" : "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations", "type" : "article-journal", "volume" : "51" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[1]", "plainTextFormattedCitation" : "[1]", "previouslyFormattedCitation" : "[6]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[1], we used a series of Sobel tests to determine the significance of the indirect effects depicted in our model. Because threat severity and threat susceptibility did not demonstrate significant direct effects on intention, neither can significantly mediate the relationship between relatedness and intention. Similarly, self-efficacy cannot mediate the relationship between competence and intention, and response cost cannot mediate the relationship between autonomy and intention due to a lack of significant direct effects from the mediators. Relatedness indirectly influenced intention through motivation (p < .01) but did not have a significant direct effect on intention, demonstrating that motivation fully mediates this relationship. Competence had an indirect effect on intention through motivation (p < .05) but also significantly affected intention directly, meaning that motivation partially mediated this relationship. Autonomy had a significant indirect effect on intention through response efficacy (p < .05) and motivation (p < .001). Because autonomy demonstrated a significant direct effect on intention as well, response efficacy and motivation both partially mediated this relationship. A summary of the mediation tests is shown in REF _Ref476840139 \h \* MERGEFORMAT Table E1.Table E1: Mediation Testing for Indirect EffectsRelationship(IVMVDV)β (IVMV)SE (IVMV)β (MVDV)SE (MVDV)t-ValueP-ValueMediation RELSEVBI0.6420.0300.0480.0341.4090.079NoneRELSUSBI0.3630.0400.0100.0310.0150.494NoneCOMPSEFBI0.2560.0470.0110.0370.2970.383NoneAUTOREFBI0.6130.0320.0880.0412.1330.016PartialAUTOCOSBI?0.3510.046?0.0240.0430.5570.289NoneRELMOTBI0.1100.0360.3190.0432.8250.002FullCOMPMOTBI0.0830.0430.3190.0431.8680.031PartialAUTOMOTBI0.5650.0340.3190.0436.7740.000Partialβ = Path Coefficient; SE = Standard Error; IV = Independent Variable; MV = Mediator Variable; DV = Dependent Variable; BI = Behavioral Intention; MOT = Motivation toward performing recommended response; REL = Perceived Relatedness; COMP = Perceived Competence; AUTO = Perceived Autonomy; TSEV = Threat Severity; TSUS = Threat Susceptibility; REF = Response Efficacy; SEF = Self-efficacy; COS = Response CostPartial Least Squares Analysis of Individual Relationships for Model ComparisonTo further assess the various models, we analyzed the individual path estimates in each model. The structural models and their associated paths were tested using SmartPLS ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Ringle", "given" : "Christian Marc", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Wende", "given" : "Sven", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Will", "given" : "Alexander", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2005" ] ] }, "number" : "2.0 (beta)", "publisher" : "SmartPLS", "publisher-place" : "Hamburg, Germany", "title" : "SmartPLS", "type" : "article" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[12]", "plainTextFormattedCitation" : "[12]", "previouslyFormattedCitation" : "[62]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[12]. PLS analysis, which uses a bootstrapping resampling technique to approximate the path coefficients and the amount of variance explained in the mediating variables, has been the most commonly used statistical tool in PMT-based information security research and is therefore used in this study as well. In examining the individual relationships in the model, we began with the paths adapted from the traditional PMT model. Response efficacy (β = .498, p < .001) and response cost (β = ?.142, p < .001) each demonstrated a significant direct effect on an individual’s intention to install password management software. Threat severity (β = .038, p > .05), threat susceptibility (β = .031, p > .05), and self-efficacy (β = .052, p > .05) did not significantly influence intention. Path estimates are described further in REF _Ref444356759 \h \* MERGEFORMAT Table E2.Table E2: Path Estimates – Traditional PMT ModelIV DV (with direction)Path Coefficient (β)t-Statp-ValueSupported?TSEV BI (+).0380.930> .05NoTSUS BI (+).0310.752> .05NoREF BI (+).49813.362< .001YesSEF BI (+).0521.199> .05NoCOS BI (-)?.1423.524< .001YesIV = Independent Variable; DV = Dependent Variable; BI = Behavioral Intention; TSEV = Threat Severity; TSUS = Threat Susceptibility; REF = Response Efficacy; SEF = Self-efficacy; COS = Response CostNext, we examined the PMT model supplemented with response performance motivation. In this model, PMT’s independent variables were depicted with direct effects on both motivation and intention. Response performance motivation was also modeled as having a direct effect on intention. Response efficacy (β = .515, p < .001), self-efficacy (β = .104, p < .01), and response cost (β = ?.115, p < .01) each had a significant impact on response performance motivation; direct effects of threat severity (β = .074, p > .05) and threat susceptibility (β = .007, p > .05) on motivation were not supported. Motivation demonstrated a significant influence on intention (β = .447, p < .001), and its inclusion in the model did not change the significance of the relationships on intention previously reported for the traditional PMT model. Path estimates are shown in REF _Ref444356945 \h \* MERGEFORMAT Table E3.Table E3: Path Estimates – PMT Model with Response Performance Motivation includedIV DV (with Direction)Path Coefficient (β)t-Statp-ValueSupported?TSEV MOT (+).0741.513> .05NoTSUS MOT (+).0070.130> .05NoREF MOT (+).51514.184< .001YesSEF MOT (+).1042.442< .01YesCOS MOT (-)?.1152.664< .01YesTSEV BI (+).0030.071> .05NoTSUS BI (+).0381.038> .05NoREF BI (+).2705.800< .001YesSEF BI (+).0010.023> .05NoCOS BI (-)?.0912.331< .01YesMOT BI (+).4479.852< .001YesIV = Independent Variable; DV = Dependent Variable; BI = Behavioral Intention; TSEV = Threat Severity; TSUS = Threat Susceptibility; REF = Response Efficacy; SEF = Self-efficacy; COS = Response Cost; MOT = Motivation toward performing recommended responseNext, we examined the impact of autonomy, competence, and relatedness on intention. Embedding motivational enhancements in the appeal for password management software installation bolstered individuals’ perceptions of autonomy, competence, and relatedness, either individually or in tandem, depending on the treatment group. As predicted in the research model, individuals’ intention to perform the appeal’s recommended response increased as individuals’ perceptions of autonomy (β = .606, p < .001), competence (β = .075, p < .05), and relatedness (β = .111, p < .01) increased. This finding indicates that each self-determined appeal manipulation is individually significant in bolstering an individual’s intention to perform a recommended response related to protecting information assets. Direct relationships between autonomy, competence, and relatedness toward intention are further illustrated in REF _Ref444357530 \h \* MERGEFORMAT Table E4.Table E4: Path Estimates – Assessing Direct Effects of Self-Determined Appeal on BIIV DV (with Direction)Path Coefficient (β)t Statp ValueSupported?AUTO BI (+).60617.460< .001YesCOMP BI (+).0751.686< .05YesREL BI (+).1112.855< .01YesIV = Independent Variable; DV = Dependent Variable; BI = Behavioral Intention; REL = Perceived Relatedness; COMP = Perceived Competence; AUTO = Perceived AutonomyNext, we analyzed the SDT model with the inclusion of response performance motivation, as well as the relationships between our motivational antecedents (autonomy, competence, and relatedness) and motivation. Autonomy (β = .567, p < .001), competence (β = .077, p < .05), and relatedness (β = .088, p < .05) each had a significant positive effect on response performance motivation. Autonomy (β = .405, p < .001), relatedness (β = .098, p < .01), and response performance motivation (β = .342, p < .001) each demonstrated a significant influence on an individual’s intention to install password management software, while competence (β = .037, p > .05) was not shown to have a significant impact on intention once motivation is included in the model. Path estimates for the full SDT model are detailed in REF _Ref444357442 \h \* MERGEFORMAT Table E5.Table E5: Path Estimates – Traditional SDT Model with Direct Effects of ACR on BI IncludedIV DV (with Direction)Path Coefficient (β)t-Statp-ValueSupported?AUTO MOT (+).56713.370< .001YesCOMP MOT (+).0771.649< .05YesREL MOT (+).0881.741< .05YesAUTO BI (+).4057.152< .001YesCOMP BI (+).0370.902> .05NoREL BI (+).0982.493< .01YesMOT BI (+).3426.278< .001YesIV = Independent Variable; DV = Dependent Variable; BI = Behavioral Intention; MOT = Motivation toward performing recommended response; REL = Perceived Relatedness; COMP = Perceived Competence; AUTO = Perceived AutonomyPost-Hoc Analysis of Protection Motivation Theory’s Influence on Motivation In our final analysis, we examined an alternative model that includes direct relationships between the PMT variables and motivation in addition to the original paths. We believe, based on prior research in self-determination, that SDT’s native antecedents (relatedness, competence, and autonomy) would provide the greatest contribution toward explaining the variance in motivation. However, it is theoretically plausible for PMT’s variables to also have an effect on motivation ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "abstract" : "This article reports the first meta-analysis of the literature on protection motivation theory (Rogers, 1975, 1983; Rogers & Prentice-Dunn, 1997), a model of disease prevention and health promotion that has generated research for over two decades. The literature review included 65 relevant studies (N = approximately 30,000) that represented over 20 health issues. The mean overall effect size (d+ = .52) was of moderate magnitude. In general, increases in threat severity, threat vulnerability, response efficacy, and self-efficacy facilitated adaptive intentions or behaviors. Conversely, decreases in maladaptive response rewards and adaptive response costs increased adaptive intentions or behaviors. This held true whether the measures were based on intentions or behaviors, and suggests that PMT components may be useful for individual and community interventions.", "author" : [ { "dropping-particle" : "", "family" : "Floyd", "given" : "Donna L", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Prentice-Dunn", "given" : "Steven", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Rogers", "given" : "Ronald W", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Applied Social Psychology", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2000", "11" ] ] }, "page" : "407-429", "title" : "A Meta-Analysis of Research on Protection Motivation Theory", "type" : "article-journal", "volume" : "30" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[4]", "plainTextFormattedCitation" : "[4]", "previouslyFormattedCitation" : "[28]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[4] (see H9). SDT’s and PMT’s variables combined to explain 47.9% of the variance in individuals’ motivation to install a password manager. Competence (β = .056, p < .05) and autonomy (β = .350, p < .001) still demonstrated a significant influence on motivation with inclusion of direct paths from the PMT variables. However, relatedness no longer significantly affected motivation (β = .037, p > .05). Response efficacy (β = .261, p < .001) and response cost (β = ?.200, p < .001) each showed a significant effect on motivation, but threat severity (β = .046, p > .05), threat susceptibility (β = .038, p > .05), and self-efficacy (β = ?.019, p > .05) did not demonstrate a significant impact. Detailed results of our post-hoc analysis are shown in REF _Ref476844981 \h \* MERGEFORMAT Table E6.Table E6: Path Estimates – Post-Hoc Analysis with PMT Effects on MotivationIV DV (with Direction)Path Coefficient (β)t Statp ValueSupported?REL MOT (+)0.0370.986> .05NoCOMP MOT (+)0.0561.684< .05YesAUTO MOT (+)0.3507.355< .001YesTSEV MOT (+)0.0461.164> .05NoTSUS MOT (+)0.0381.163> .05NoREF MOT (+)0.2614.188< .001YesSEF MOT (+)?0.0190.430> .05NoCOS MOT (-)?0.2005.791< .001YesREL TSEV (+)0.64221.298< .001YesREL TSUS (+)0.3639.101< .001YesCOMP SEF (+)0.2565.397< .001YesAUTO REF (+)0.61319.059< .001YesAUTO COS (-)?0.3517.568< .001YesTSEV BI (+)0.0481.436> .05NoTSUS BI (+)0.0100.308> .05NoREF BI (+)0.0110.305> .05NoSEF BI (+)0.0882.185< .05YesCOS BI (-)?0.0240.572> .05NoMOT BI (+)0.3197.334< .001YesREL BI (+)0.0130.356> .05NoCOMP BI (+)0.1143.603> .05NoAUTO BI (+)0.3447.242< .001YesCOMP BI (+)0.0370.902> .05NoREL BI (+)0.0982.493< .01YesMOT BI (+)0.3426.278< .001YesIV = Independent Variable; DV = Dependent Variable; BI = Behavioral Intention; MOT = Motivation toward performing recommended response; REL = Perceived Relatedness; COMP = Perceived Competence; AUTO = Perceived Autonomy; TSEV = Threat Severity; TSUS = Threat Susceptibility; REF = Response Efficacy; SEF = Self-efficacy; COS = Response CostReferencesADDIN Mendeley Bibliography CSL_BIBLIOGRAPHY 1.Baron, R.M. and Kenny, D.A. The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations. Journal of Personality and Social Psychology, 51, 6 (December 1986), 1173–1182.2.Churchill, G.A. A Paradigm for Developing Better Measures of Marketing Constructs. Journal of Marketing Research, 16, February (1979), 64–73.3.Crump, M.J.C., McDonnell, J. V, and Gureckis, T.M. Evaluating Amazon’s Mechanical Turk as a Tool for Experimental Behavioral Research. PLoS ONE, 8, 3 (2013), e57410.4.Floyd, D.L., Prentice-Dunn, S., and Rogers, R.W. A Meta-Analysis of Research on Protection Motivation Theory. Journal of Applied Social Psychology, 30, 2 (November 2000), 407–429.5.Gefen, D. and Straub, D.W. A Practical Guide to Factorial Validity Using PLS-Graph: Tutorial and Annotated Example. Communications of the Association for Information Systems, 16, (2005), 91–109.6.Loch, K.D., Straub, D.W., and Kamel, S. Diffusing the Internet in the Arab World: The Role of Social Norms and Technological Culturation. IEEE Transactions on Engineering Management, 50, 1 (February 2003), 45–63.7.Lowry, P.B. and Gaskin, J. Partial least squares (PLS) structural equation modeling (SEM) for building and testing behavioral causal theory: When to choose it and how to use it. IEEE Transactions on Professional Communication, 57, 2 (2014), 123–146.8.Mackenzie, S.B., Podsakoff, P.M., and Podsakoff, N.P. Construct Measurement and Validation Procedures in MIS and Behavioral Research: Integrating New and Existing Techniques. MIS Quarterly, 35, 2 (2011), 293–334.9.Peter, J.P. Construct Validity: A Review of Basic Issues and Marketing Practices. Journal of Marketing Research, 18, 2 (1981), 133–145.10.Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., and Podsakoff, N.P. Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies. Journal of Applied Psychology, 88, 5 (October 2003), 879–903.11.Podsakoff, P.M., MacKenzie, S.B., and Podsakoff, N.P. Sources of method bias in social science research and recommendations on how to control it. Annual Review of Psychology, 63, (January 2012), 539–569.12.Ringle, C.M., Wende, S., and Will, A. SmartPLS. 2005. , Z.R., Hammer, B.I., and Limayem, M. Data Collection in the Digital Age: Innovative Alterantives to Student Samples. MIS Quarterly, 38, 2 (2014), 355–378. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download