By Kyle Lundby, Ph.D., Jeffrey A. Jolton, Ph.D. and Siddartha Parthasarathy, Kenexa

Employee survey programs can provide organizations with the unique and invaluable opportunity to link human capital metrics to survey results. This linkage may simply provide specific information about where one works in the organization and who they report to, contain more detailed demographic information (e.g., gender, length of service) or indicate specific performance data (e.g., performance evaluation scores, high potential membership). Regardless of what information is linked to the survey, this opportunity requires that the survey itself provide some identifying information about the individual taking the survey so it can be merged with other non-survey metrics.

However, trust in the process becomes crucial to the success of the survey program. Several standards can be put in place throughout the survey process to assure an appropriate level of anonymity and confidentiality. Concerns about confidentiality also have to do with organizational history/culture and employee subjective perceptions. Generally, employees feel less comfortable as the number of personal and organizational demographic questions increases.

Organizations now more frequently consider attributed surveys as a tool for improving data quality and their ability to link to other outcomes. Attributed surveys have this information embedded with the survey process that is provided to Kenexa by the client organization. However, attributed surveys can create confidentiality concerns, which are less potent in non-attributed surveys.

Non-attributed surveys are those where no individual identifying  information about the participant is embedded and the participant  directly shares the information on the survey.

There are advantages and disadvantages to attributed and  non-attributed surveys, some of which are summarized here.

Advantages of Attributed Surveys

Insight analyses: By linking other data (e.g., HR metrics) to  employee opinion survey data, additional analyses can be run to demonstrate linkages between these metrics (e.g., employee  opinions in particular workgroups and how they related to  manager evaluations, customer ratings, financial ratings). This  provides a deeper level of insight from the results and explains how  elements measured in the survey drive or relate to other business  metrics. For example, one Kenexa financial services client with a  fully attributed survey has been able to combine actual employee  turnover (obtained through their HR database) with its employee  opinion survey results. By combining these two different sets of  data, Kenexa researchers were able to develop a predictive model  of actual turnover so that actions could be put in place to reduce  future turnover. 

Shorter survey: Key demographic information that is used to create reports (e.g., department, location, function) can be provided to Kenexa and linked to the survey results. Consequently, these demographic questions do not need to be asked on the survey. While space savings will depend on the total number of demographics of interest, it is not unreasonable to expect to reduce the length of a survey by one or two pages. This helps conserve time (because employees do not need to respond to those questions) and paper.

Data accuracy: When allowed to self-identify demographics, there is a risk that individuals will miscode themselves. However, with group identifiers and pre-loading of information, we can minimize this risk. This is especially valuable in extremely large organizations (five to six levels of hierarchy) or where organizations have gone through substantial or continual restructuring or mergers.

Accurate response rates: Related to data accuracy, response rates are likely to be more accurate with attributed surveys (non-attributed surveys in which individuals self-code their workgroup can exhibit response rates in excess of 100%).

Disadvantages of Attributed Surveys

Loss of control: Individuals sometimes choose not to respond to demographic questions for various reasons. By pre-loading such information, employees are no longer able to decline such questions. This can be managed somewhat by limiting the information that is pre-loaded. In some cases, employees are given the opportunity to “opt out” of the process—allowing them to continue taking the survey without the information linked later.

Fear: Regardless of actual intent and pre-communication about confidentiality and anonymity, some employees may be suspicious of the motives behind an attributed survey approach. Transparency in the process, such as providing clarity about what is being attributed and what will be done with the information, helps to mitigate some of this fear. Furthermore, employees find reassurance if the identifying information and research are being handled by the vendor and not the client organization.

Declining response rates and/or less candid responses: Because of the disadvantages cited above, some individuals may choose not to complete the survey, which could harm response rates. Moreover, survey participants may not be as forthcoming with their feedback, especially negative feedback, if they feel their results lack anonymity.

Advantages of Non-Attributed Surveys

Individual control: By asking individuals to enter their workgroup membership and/or their personal demographic information, they can feel more control over what they do and do not share with the organization.

Simplicity: Pre-loading information requires an up-to-date HR database. If that is not the case, workgroup membership based on the pre-loaded information may actually be less accurate than if individuals were asked to enter that information on the survey.

Disadvantages of Non-Attributed Surveys

Incorrect classifications: For various reasons, individuals often miscode their workgroup membership. One consequence is that manager reports can contain data for individuals that do not belong in their groups.

Response rate problems: Another consequence of incorrect classifications is that response rates may be inflated for some groups. This is particularly problematic, for example, when a particular workgroup obtains a response rate in excess of 100%.

Unanswered questions: Many questions asked by senior leadership can only be answered when conducting attributed surveys. The specific link of results to individual performance and behaviors (e.g., absenteeism, performance, high-potential groups) is often of interest to leadership, but cannot be addressed with nonattributed surveys.

Conclusions

Choosing one approach or the other—attributed or nonattributed—depends on a number of factors, including history, culture, quality of the current HRIS database and the desire for post-survey analytics. All of these factors must be considered to arrive at an appropriate decision—one that will maximize survey returns, minimize concerns of employee stakeholders and allow for any follow-up analytics.

However, this decision does not need to be mutually exclusive. For some organizations, a middle ground may be the best solution. For example, those organizations that are not ready for a fully attributed survey may opt for pre-loading some information (e.g., region, location) while leaving the rest up to employees (e.g., personal demographics such as age, gender). This approach has the dual advantage of mitigating confidentiality concerns, while allowing for linkage research at a higher (e.g., store location) level. In one example, a Kenexa client in the food retail industry pre-coded surveys to store locations. At this level, employees were less likely to be concerned that they might be personally identifiable and it also allowed Kenexa to combine these survey results with store-level Market Research (e.g., customer satisfaction at the store level) and HR metrics (e.g., employee turnover at the store level) to create a model showing the drivers of customer satisfaction and employee turnover for that organization.

Whatever direction organizations may choose, trust is always a critical component of a successful survey program. Trust is often bound by the degree of anonymity and confidentiality a survey provides. Several standards need to be put in place throughout the survey process to assure an appropriate level of anonymity and confidentiality. Kenexa consultants are well versed in the pros and cons of attributed and non-attributed surveys and can work with client organizations to address any issues of trust that may potentially arise. This allows companies to choose the best solution that works for them.

About the Authors

Kyle Lundby, Ph.D., is a senior global survey consultant for Kenexa. He recently returned to the United States after serving as the lead consultant for the Asia Pacific region. He has more than 16 years of experience working with organizations in the private and public sector on all phases of employee and customer stakeholder measurement and consulting. While in Asia, he worked with organizations representing a variety of industries, including financial services, insurance, retail, consumer products and energy. In addition to working with Kenexa clients in Hong Kong, he frequently traveled to China, Japan, Malaysia, Singapore, India and Australia. In his current role, Dr. Lundby partners with clients to implement and interpret the results of stakeholder measurement systems, including employee engagement surveys, customer loyalty tracking studies and multirater feedback surveys.

Before joining Kenexa, Dr. Lundby held executive consulting and director positions with other leading research firms. He is a longtime, active member of the Society for Industrial and Organizational Psychology (SIOP) and sits on the editorial board of SIOP. Dr. Lundby is also editing an upcoming (Spring 2010 release) SIOP Professional Practice Series book: “Going Global: Practical Applications and Recommendations for HR and OD Professionals in the Global Workplace.” Dr. Lundby is a frequent presenter at conferences and has published numerous journal articles and book chapters on employee engagement, linkage research and methodologies to demonstrate the return on investment of HR initiatives. He holds a Doctorate degree in organizational psychology from the University of Tennessee and a Master degree from San Diego State University

 Jeffrey A. Jolton, Ph.D., is the director of global services at Kenexa, overseeing the project management and thought leadership for many of Kenexa’s largest global survey projects. Dr. Jolton has over 15 years of extensive consulting experience, which have made him a leading expert in assessment-based organizational change. He has applied his innovative and insightful guidance to number of global businesses ranging from retail giants such as The Home Depot and Gap Inc. to leading financial and professional companies such as Allianz, HSBC, PricewaterhouseCoopers and Wachovia, as well as other global organizations such as DPWN, Textron, and Pitney Bowes. As Director of Global Services, Dr. Jolton oversees project teams in all corners of the world including Germany, Hong Kong, Singapore, the United Kingdom and the United States.

Dr. Jolton is a regular presenter at numerous professional conferences and has more than 25 publications in professional and scientific journals. Recent topics of his works include understanding dysfunctional organizations, global truths about employee engagement, the role of critical thinking in strategic planning and behavior change across the global spectrum. He holds a Doctorate and a Master of Science degree in Industrial and Organizational Psychology from Ohio University.

Siddartha Parthasarathy is the director of client services for Kenexa’s Asia Pacific region. He currently leads the Singapore office and helps grow the APAC region by supporting global and regional accounts. He has had extensive experience in employee consulting, directing complex global projects and providing consulting advice to the financial, pharmaceutical, telecom and manufacturing sectors. Partha began his career as a strategy consultant. He holds a degree from the Wharton School of Business.