Personality and software developers’ job performance

Bullet points:

• personality traits can account for over 40% of software developers’ job performance;
• conscientiousness and emotional intelligence have the greatest impact on performance;
• IT companies can employ a scientific approach to improve HR and management processes.

The misconception that IT specialists in general are less likely to show personality traits related to social interactions and as a result might have poorly developed social skills, has been around for a long time. This conviction, insufficiently researched, impacts heavily HR processes related to IT staff, in particular recruitment. We believed that the role of non-technical, so-called soft skills (such as personal characteristics, i.e. personality traits) is underestimated when it comes to IT staff management, and during the study to be discussed we have found out that this belief is shared by many IT professionals as well.

In this article we would like to present how to employ a scientific approach to address such a myth and gather data that could help improve HR procedures in IT. Eventually, we would like to present evidence that IT companies might benefit from including personality traits (often referred to as ‘soft skills’ in every-day communication) into HR processes.

As researchers interested in HR, we decided to investigate the actual impact that personality has on work-related behaviors in IT, with particular emphasis on job performance. Based on known evidence and scientific theories, we expected this impact to be substantial. Personality can improve the developer’s performance by shaping every-day work-related habits. A conscientious developer may be more alert to his or her work outcomes, which may result in fewer mistakes. A more open-minded individual may be more eager to master a new skill or technology and consequently he or she may deliver the desired solution in an effortless or more error-proof manner. Besides, personality can make it all the easier for developers to behave appropriately and as required in a specific situation and context. For example, developers who recognize and cope more efficiently with their and others’ emotions  may find it easier to respond adequately in a highly demanding environment, where interpersonal conflicts resulting from stress and strain can easily occur. This will be the case especially when communication is handled remotely, via mail, IM or other technology, which hinders decoding the other person’s motive (which in turn, is kind of bread and butter for IT). These are our hypotheses. If they were to be confirmed in real-life, business context, it could mean that personality measures (such as questionnaires or structured interviews) could considerably enhance the validity of recruitment processes.

What did we do to investigate this? We chose reliable tests which measure personality traits and we designed tool to estimate individual job. Then we invited Polish software developers to fill in our survey. Over 220 people responded and provided us with estimations of both their personality and job performance. Then we took a hard look at the data gathered to determine how our hypothesis corresponds to reality.

What did we find out? Personality alone can explain over 40% of job performance of software developers! Is this a lot? Well, if you are used to predictive models built on financial or traffic data, that may not impress you at first. However, human behavior is not that easy to measure and predict, and in social sciences we rarely this numbers to be greater than 30%. Besides, mind that we measured personality traits solely as performance antecedents. We did not take technical skills, experience or motivation under consideration, nor any other individual or external factor that leads to better or worse performance. Put that way, in our humble opinion, these numbers demonstrate how substantially influential personality is in the IT job context.

Are these results robust? We employed a cross-validation procedure to find out, and it was successful. However, to determine that further, we carried out a second study to test our results, in which we employed identical tools and methodology. This time we gathered a smaller but international group of participants (from all over the world). We used the machine learning approach and checked whether the exact model from the first study allows us to predict job performance levels of 60 new participants. And it did!

Now that we have provided evidence of the significant role of personality in IT jobs, there is one more question to be asked: “Which personality traits are actually the most important ones?” Step regression results together with additional relative weight analysis showed that two traits contributed almost exclusively to the prediction of job performance: conscientiousness and emotional intelligence (this might be surprising, but in fact emotional “intelligence” is considered by modern psychologists simply as a personality trait related to emotion perception and social behaviors). Further analysis with the use of clustering also confirmed that only when these two traits were evaluated highly, was the performance also evaluated as satisfactory (this one part of results, however, should be treated only as a preliminary conclusion, as sample size was not sufficient to determine stable sub-groups of participants).

Finally, how can these results be applied to the business organization context? First, if you are a developer, you can utilize them if you are seeking to improve your work. There are of course many ways to achieve exceptional performance, and if you already have found your own approach, stick to it! However, the results might be worth giving a shot. As personality is nothing else but behaviors manifested in certain situations, our study findings imply that it might be beneficial to focus on particular behaviors related to conscientiousness and emotional intelligence (such as maintaining self-discipline, performing in a planned manner, communicating your concerns freely, controlling your emotional responses). We have provided evidence that these behaviors (among many others), if manifested, could lead to enhanced performance.

Secondly, if you are a team leader, HR practitioner or manager, it can be beneficial to your company to employ personality measures into your HR procedures. Include personality traits into your desired candidate profile, especially those related to interpersonal behaviors. As personality is quite stable for a given individual over an extended period of time, the simple-to-gather information about one’s personality may help you to predict how well this person will fit your company and how well they will perform.

Lastly, there is more! What we have done is neither excessively expensive, nor time-consuming (but of course it does require some level of knowledge and methodology skills). Our results are quite robust, but they concern developers in general. If you are interested in evidence-based management and you seek to enhance your company’s HR procedures, there is no reason why you shouldn’t carry out a similar study in your own company. This may result in obtaining a more custom-made and sophisticated model that will prompt which personality traits (or other significant personal characteristics) should be included into your HR practices. Such a study could increase the validity of personnel decision taken in your company with respect to your company culture and specific nature. Researchers like us will be happy to help with such studies!

Jaros?aw Grobelny

Sonia Chojnik-Konopa

Aneta Ja?lanek