ITMPI FLAT 003
Main Menu
Home / IT Best Practices / The 5 Leadership Styles for Managing Data Scientists

The 5 Leadership Styles for Managing Data Scientists

Inappropriate leadership is the root of why most data science experiments fail. John Weathington, writing for TechRepublic, explores why projects fail because of unfitting leadership and how to choose a management style that is effective for addressing different problems. According to contingency theory, the leadership style should adjust according to the situation. Victor Vroom and Phillip Yetton developed this model and suggest five different decision-making styles, dependent upon the amount of group involvement:

  1. Autocratic leadership
  2. Modified autocratic leadership
  3. Consultative leadership
  4. Modified consultative leadership
  5. Group-based leadership

Dictate or Collaborate

Autocratic leadership simply put is to make all the decisions yourself without any input from outside sources. Although this method is not highly encouraged, there is a time and a place that makes this effective. The key element to remember when choosing this style is that speed is important, being both quick and efficient.

Similar to autocratic leadership, the “modified” version implements outside knowledge that you are not aware of. The trick is to ask very direct questions to the data scientists, but you do not let on as to why you are asking. However, there is potential damage control that may need to be implemented. The data scientists may feel used, because they sense something going on that you are unwilling to share.

When it comes to consultative leadership, rather than making independent decisions, you would involve the data scientists. This method allows for a different insight from an expert on the subject. The drawback to this sort of style is that decisions may take time because you need to talk to each data scientist individually, allow them to process the information, and then have another conversation about what to ultimately do.

Instead of separately visiting each data scientist and asking for their opinion, you could just hold a group meeting to discuss the problem. This is modified consultative leadership. It is imperative however that you are up front that, no matter the consensus of the group, the decision is ultimately yours and will be executed as you see fit.

Finally, there is group-based leadership. In this scenario, the team makes the decision, a far opposition to an autocratic style. If this style is used too frequently, the group may begin to question your assertiveness. Do not over-delegate your authority onto the group; it is important to remain assertive and decisive on your own.

Leading a team can be an arduous, overwhelming task, but there is always a good way to handle decision-making. Choose a style that makes the most sense, will be the most efficient for the scenario, and will produce the best outcome.

You can read the original article here: http://www.techrepublic.com/article/5-effective-leadership-styles-for-managing-data-scientists/

About Danielle Koehler

Danielle is a staff writer for CAI's Accelerating IT Success. She has degrees in English and human resource management from Shippensburg University.

Check Also

How to Work from Home When You Have Kids

Working from home may seem like a dream come true for a parent. You get …

Leave a Reply

Your email address will not be published. Required fields are marked *