Realizing the effects of trust and personality in cross functional teams using ANFIS classification framework

Abstract

Social behaviors are an integral part of team building. In this context, we propose a novel classification model that chooses an optimal classifier from the pool of classifiers for predicting the overall performance (OP). Secondly, the chosen classifier is used to investigate the impact of trust and personality on OP. To achieve these goals a pilot study with real time data from 442 respondents are collected from cross functional teams (CFTs) in India using an E-Questionnaire system. The results indicate that the adaptive neuro fuzzy inference system (ANFIS) method is an optimal classifier (A = 89.14%) with respect to other classifiers. We also infer that the predictors, trust and personality are most suitable for predicting OP with a direct relationship to OP and play an indispensable role; as a catalyst; for boosting OP.

Appendix

1.1 Trust and personality factors

The trust and personality are the two essential parameters for determining the overall performance of the project. This overall performance is the combination of both the internal and external performance. Before we formulate the factors for trust and personality, we define the two terms. The word trust whose origin is from a word “traust” (Old Norse) which means strong, is defined as “the belief that someone or something is reliable, honest, good and effective” as given by Merriam Webster. The word personality whose origin is from the word “personalis” (French word) which means character, is defined as “the set of behavioral and emotional qualities that differentiates one person from another” as given by Merriam Webster.

The trust and personality factors are formulated into a set of five likert scale questionnaires. The set of factors are derived from Zhang and Zhang (2015) and Goldberg (1992) respectively. These questions are framed as an online feedback system and are uploaded in the website “www.funbrainquest.com ”. For clear understanding about the set of questions regarding trust and personality refer [16, 60].

1.2 Description on trust and personality factors

1.2.1 CS factors

Team diversity It describes the unique nature of an individual in a team. It involves age, ethnicity, gender etc.

Team longevity It describes the migration of members from a team. The dynamics of the team is investigated in team longevity.

Team proximity This describes the demographic location of the team members. It deals with whether the members are closely located or separated far. The members are either face–face interacting or virtually interacting

Justice measure This deals with the typical nature of how meetings were conducted in different firms and how protocols were initiated in these firms. Here, procedural and interactional justice is being concentrated.

1.2.2 PP factors

Affective ManieBosman defines this as a form of trust developed as a result of the emotional blend between members of a team. It is based on feelings and intuitions with no rational reasons.

Cognitive Manie Bosman defines this as a form of trust which has a rational reason. This trust develops when one considers someone or something as reliable, competent and responsible. These characteristics give a rational reason for the trust.

Cooperative Costa and Anderson define this form of trust as an act of showing willingness to be vulnerable for others actions. It deals with the aspects of integrity, communication etc.

1.2.3 Personality factors

Openness It refers to people who are open minded, curious, willing and emotionally stable to take up critical tasks.

Consciousness This refers to people who are self disciplined, honest, loyal at work and aim for higher scores of achievements.

Extraversion These are people with higher rates of enthusiasm. They are focused, like interactions, like getting along with diversified set of people and highly energetic.

Agreeableness It refers to people who get along easily with other people. They are kind, caring, affectionate and optimistic in their view points.

Neurotics This personality trait is a deionizer. It deals with negative emotions like ego, stress, anger, frustration, hatred etc. People with this personality tend to take things emotionally negative and are highly reactive.

1.3 Performance metrics

There are several metrics used to estimate the feasible performance of a classifier. These metrics originate from the confusion matrix which is a matrix formed between the actuals and predicted set of class instances. The key metrics used for estimating the classifier’s performance are sensitivity, specificity, accuracy and kappa coefficient.

Sensitivity It is also called true positive rate or recall. This is the measure of proportion of positives correctly classified. It is given by $\left( {\frac{truepositive}{positive}} \right)$ where positive is the sum of true positive and false negative.

Specificity It is also called true negative rate. This is the measure of proportion of negatives correctly classified. It is given by $\left( {\frac{truenegative}{negative}} \right)$ where negative is the sum of false positive and true negative.

Accuracy The ratio of sum of true positives and true negatives to total number of instances. It is given by $sensitivity\left( {\frac{positive}{total}} \right) + specificity\left( {\frac{negative}{total}} \right)$ where total is given by sum of true positive, true negative, false positive, false negative.

Kappa coefficient It is a powerful statistical measure which evaluates the inter-rater agreement. It is better than accuracy measure as it takes occurrence by chance also into consideration. It is effective for qualitative data. The kappa coefficient is calculated using $\left( {\frac{{p_{{o - p_{r} }} }}{{1 - p_{r} }}} \right)$ where $p_{o}$ is the accuracy or observed probability and $p_{r}$ is the random probability.