Merging Explicit Declarations With Implicit Response Time to Better Predict Behavior

Merging Explicit Declarations With Implicit Response Time to Better Predict Behavior

Rafal Ohme (Stellenbosch University, South Africa & NEUROHM, Poland), Michał Matukin (NEUROHM, Poland & Walnut Unlimited, UK) and Paula Wicher (NEUROHM, Poland)
DOI: 10.4018/978-1-7998-3115-0.ch023
OnDemand PDF Download:
No Current Special Offers


Declarations and actions do not always overlap, and thus, predicting future behavior solely on the basis of self-reported measures seems to be ineffective. The authors propose a confidence index (CI): a measure based on Fazio's attitude accessibility model. CI integrates explicit and implicit perspectives and captures how long a person hesitates when stating an opinion. The more certain someone is the stronger the attitude-behavior link is likely to be. A study was conducted to uncover differences in attitudes between average- and top-performing sales agents from the automotive industry. The results for declarative data did not show any significant differences; however, the CI results yielded interesting significant differences between groups. Random decision forests analyses confirmed that merging explicit and implicit measures increases predictive power of the tool. The study provided actionable insights on how to improve sales team performance, which were then implemented and eventually validated by sales results.
Chapter Preview


Attitudes reflect people’s experiences and perceptions. Most often they are defined as ‘mental states regarding an object or issue’ (Cooper, Blackman, & Keller, 2016). They provide a strategy for evaluating people or events, they involve a complex organization and guidance of evaluative beliefs, memories, and self-esteem. Attitudes are broadly applied and often used to analyze and interpret such phenomena as self-concepts, social influence, interpersonal attraction, social support, political and religious views or consumer and organizational behavior (Hussak & Cimpian, 2018; Prislin & Wood, 2005; Pornsakulvanich, 2017; Abdalla, Abdelal, & Soon, 2019; Alves, 2018; Alhawari, Verhoff, Ackermann, & Parzeller, 2019; Ajzen, 2008; Shamim, Cang, & Yu, 2019). Attitudes have also been associated with having the capacity to influence behavior (Fishbein & Ajzen, 1975; Ajzen, 1991; Ogunnaike, Kehinde, Omoyayi, Popoola & Amoruwa, 2017).

Researchers had already made the first attempts to structurally and systematically measure attitudes already in the early part of the 20th century. They did it by asking explicit questions and assigning numbers to the given answers in order to be able to compare them between respondents (Kruglanski & Stroebe, 2012, Cooper, Blackman, & Keller, 2016). Thurstone (1928) was the first author of an attitude scale measurement, which consisted of statements concerning a particular issue with a numerical value indicating how favorable or unfavorable it was judged to be. The respondents’ task was to choose which statement they endorsed with the use of “agree” and disagree” answers. Likert (1932) suggested agreeing or disagreeing with the given statement at various strengths on the 5-point scale. Osgood (1962) proposed that attitude could be measured on a 7-point scale along three dimensions: evaluation (good/bad), potency (strong/weak), activity (active/passive). Attitudes were also measured with the use of semantic differential scales, which was a set of evaluative bipolar scales (Triandis & Fishbein, 1963).

One of the questions that emerged from a traditional or explicit approach to attitude questionnaires is how can researchers be certain that people declare their actual point of view? Thurstone himself stated that attitude scale usage should be limited to situations where researchers expected participants to tell the truth about their opinions (Thurstone, 1928). Tapping into real motives of human behavior has always been a challenge. Declarations and actions do not always overlap. A good example of such inconsistency one can be observed during political elections. People declare they will vote for a specific candidate, however, on the election day the results are not always in line with the ones predicted by the polls. This happens, because some motives of people’s behaviors are not consciously accessible. In the last 50 years researchers have consistently shown that most of our cognitive processes take place outside of conscious awareness and control, nevertheless they do influence our perceptions, judgments and actions (Zajonc, 1968, 1980, 1984; Holender, 1986; Uleman & Bargh 1989; Bornstein 1989; Bornstein & Pittman 1992; Greenwald 1992; Murphy & Zajonc 1993; Murphy, Monahan & Zajonc 1995; Niedenthal & Kitayama, 1994; Bargh 1994, 1997; Ohme, 1997, 2001, 2009; Nosek, Hawkins, & Frazier, 2011; Ohme, Pochwatko & Blaszczak, 1999; Ohme & Boshoff, 2019; Matukin, & Ohme 2016, 2017). Especially in socially sensitive contexts, people may declare attitudes that are expected and approved. However, attitudes measured by explicit tools are based on rationally processed information, complying with social norms, evaluation apprehension or based upon “cost-benefit analysis” (Dovidio & Fazio, 1992).

Key Terms in this Chapter

Attitude-Behavior Link: A theory that explains conditions when attitudes influence behavior. The strength (the confidence) with which an attitude is held is a good predictor of behavior and can be assessed thanks to the response time measurement.

Response Time: The time our mental system needs to process information, it can be used as a measure of the time course of mental operations in the human nervous system. As a theoretical background has been continuously used and developed since late 1800. It stormed the academic world and was popularized thanks to application in such approaches as Priming or Implicit Association Test.

Implicit: (a) Implicit or unconscious processes - traces of past experience that are unavailable to self-report or introspection, e.g. stereotypes; (b) Implicit or indirect measures – approaches used to enhance understanding of attitude formation, associations, strength and accessibility, especially when respondents are unable or unwilling to name the sources of influence on their behavior or opinion. During the testing procedure respondents are unaware of what is examined (e.g. stereotypes) and how it is examined (e.g. by measuring response time).

Confidence: An attitude related feature that is connected with such phenomena as attitude strength or attitude accessibility. The higher the confidence the stronger and more accessible the attitude will be. Confidence can be measured with implicit tools that are utilizing response time measurement. It is directly linked with these aspects of attitudes that drive behaviour and can be used as an indicator of the strength of the attitude-behaviour relationship.

Confidence Index: A standardized index in which declarative responses are weighted by their response time. It reflects how confident an attitude is. This merge of explicit and implicit measures enabled more advanced statistical computations. Confidence Index works in a similar way to a security ribbon in a banknote and indicates whether the answer is authentic and valuable or false and abused.'

Attitude: A set of emotions, beliefs, and behaviors toward a particular object, idea, person, thing, or event and can have either a positive or negative valence. There are two kinds of attitudes: explicit attitudes that are conscious beliefs and implicit attitudes that are unavailable to self-report or introspection.

Complete Chapter List

Search this Book: