Understanding the human elements and how to work with them are important aspects of data science. The field of organizational psychology has given us many insights into human behavior, motivations, fears, and values that can help data scientists understand their customers better. This blog post will cover four reasons why every data scientist should study organizational psychology to make more informed decisions about their products or services in order to create something that is not only effective but also appealing.
1. Data scientists are not just statisticians —we must also be able to understand people
Let’s face it, it easy to add a bunch of input data and see if any of them are good predictors for supervised models or a nice set of clusters for unsupervised models. Once we can get the model to an “acceptable” level, we often proceed with interpretation of the results based on people’s (e.g. customers) behavior and motivation. However, there are times when we use psychology to explain our model retrospectively may not be the preferred way.
We are data scientists because we have the technical skills to build a product or answer business questions using data even if that data is incomplete and contains errors. However, when we are trying to predict an outcome that involves people’s behavior and motivation, data scientists need to be able to understand what factors can influence their decisions, preferably before start collecting data and testing.
In other words, data scientists should not just use data when building a product but also consider psychology in order to create something truly effective for customers. This is where organizational psychology comes into play as it has many insights into human nature around working, decision-making as well as motivation and fears.
Organizational psychology has given data scientists a better understanding of how and why people make certain decisions or behave in specific ways that can help data scientists create something informative for their customers rather than just another product that might have missed the mark.
2. Studying organizational psychology will introduce you to concepts like social cognition, and self-perception theory which are applicable across all industries
For example, social cognitive theory explains the relationship between people, environment, and behavior. This theory is commonly used to develop programs to initiate behavioral changes. For data scientists, when designing a model that requires an intervention of some sort to trigger a behavioral change, having a good understanding this theory would make a significant impact on your approach.
Self-perception theory is another example of a common term that has its roots in organizational psychology. This theory states that people tend to make inferences about their own thoughts, feelings, or intentions based on the actions they take rather than observing their actual behavior directly. This means that data scientists must carefully design data collection to get the most accurate information about behavior or activity. In addition, capturing perceptual variables can also be a critical part of this process.
3. Studying organizational psychology will teach you how to be more aware of your own biases and assumptions when analyzing data
These organizational psychology concepts will help data scientists understand how people think and make decisions. For example, we should be aware about cognitive biases which can affect data science projects in big ways (and that we are all prone to). For example, confirmation bias is when we tend to look for information that confirms our preconceptions and ignore data points that don’t match up with those preconceptions. I am sure we are aware of this bias but it is hard to keep in check.
Another type of bias data scientists should be aware of is the anchoring bias. This happens when data points are over-influenced by data we looked at earlier in a decision making process (for example, what data to use for our analysis). There’s also groupthink which occurs when everyone agrees with an idea and nobody thinks critically about it or brings up opposing perspectives can be harmful. All data scientists should be aware of these biases and how they can affect data science projects.
4. Organizational psychologists study topics such as leadership styles, group dynamics, motivation, and conflict resolution — all of which are important for any data scientist looking to build successful teams or lead others
As data scientists we need to understand the psychology of our data sets in order to work with data effectively. We also need to motivate ourselves and others so that everyone is doing what it takes to deliver results on time and under budget. You might be a team leader or an executive that lead a data science team. There are many data science roles that require someone to lead others. If you are a data scientist in this role, understanding the psychology of your data scientists is essential for success as a team leader and executive.
Organizational psychologists study topics such as leadership styles, group dynamics, motivation, and conflict resolution — all of which are important for any data scientist looking to lead a team. Setting well defined goals that your direct reports understand and allowing them to take ownership of their work are examples of strong leadership. Thus having a deeper understanding these psychology based concepts and putting them to use for your daily work would result in much more productive and having a more fulfilling work experience for you and the team.
I hope this article has given you some new ideas on how to think about organizational psychology in relation to your data science work. It’s a topic that can seam obvious, but one thing is for sure: the more knowledge and perspective we have as data scientists, the better our work will become. So if you are wondering why every data scientist should study Organizational Psychology or want help understanding what it means for your company — leave comments below. I am happy to connect with anyone who wants continue discussions on this topic.
Relating to this article, I also published an article on the importance of stakeholder management for data science projects here. Feel free to let me know if you like these types of articles as well.