The Consequences of Mistakes in Human Services Decision Making

Our Staff

NCCD Blog

The Consequences of Mistakes in Human Services Decision Making

Kristen Johnson, PhD, Senior Researcher

When assessing risk—especially relative risk (the risk of one outcome happening versus another)—our cognitive biases often get in the way. For example, some of us worry about airplane crashes, when the risk of being in a car accident actually is much higher. Even if I am anxious about flying, I would be better off ensuring my safety while in a car: reaffirming my will not to text and drive and remembering to have the car inspected. In other words, a correct assessment of risk can help me engage in preventive steps to reduce my individual risk of more-likely threats of harm.  

Predictive analytics and research-based risk assessments help support decisions by producing a research-based risk classification to help compensate for our cognitive biases. A valid, reliable, and equitable risk assessment tool informs decisions as varied as: 

  • How much to charge for car insurance;
  • Whether to recommend preventive steps to reduce one’s risk of heart attack;
  • Whether to recommend a full mastectomy in order to reduce the risk of future breast cancer;
  • Whether to recommend intensive services for a youth at high risk of future juvenile justice involvement;
  • Inpatient versus outpatient mental health treatment; and
  • Whether someone should be released for parole or not, and if released, how intensively to serve and supervise that individual.

Two types of mistakes are possible in such decision making: false positives and false negatives. A false positive is similar to raising a red flag erroneously, whereas a false negative occurs when the flag was not raised but should have been. The implications of mistakes in these risk-informed decisions for civil liberties and individual health vary immensely.  

In the case of car insurance risk estimation, a false positive case is one in which someone (me) is classified as “high risk” and therefore charged a higher premium, but I never submit a claim nor have an accident. The consequences for individual health and civil liberties in this example are minimal: I pay a higher-than-necessary premium, which may have a slight negative impact on well-being, but this does not adversely affect my health or civil liberties. Driving is a choice, especially when other means of transportation are available. On the other hand, consider a case of denying parole to someone who is classified as being at high risk of subsequent criminal behavior but, if released, would not actually commit another crime. The consequences of a false positive case for civil liberties and individual health are much higher. 

We must weigh the consequences of mistakes when deciding how to best use risk assessment and predictive analytics information. Imagine a child protection agency using a risk assessment to classify individuals and families by the likelihood of child maltreatment. The resulting classification could be used to inform prevention-oriented, evidence-based programming, or it could be used to remove a child from his or her home. A false positive case for the latter decision has high costs regarding civil liberties and individual health, especially when you consider that being in foster care is correlated with numerous negative outcomes for children. Thus, the more rational choice is to provide effective, evidence-based programming to families identified as being at high risk of future child maltreatment.  

Ensuring that risk assessments are valid, reliable, and equitable is critical. Of equal importance is examining how risk assessment information is used and to ensure that people’s civil liberties and health are protected in the case of mistakes. In social services, it can be very effective to combine a valid, reliable, and equitable risk assessment with prevention-oriented, evidence-based practices to improve outcomes for individuals. Numerous child welfare and juvenile justice agencies are doing just this. See, for example, Wyandotte County, Kansas, which uses a risk assessment and structure to target multi-systemic therapy services to youth at high risk of future delinquent acts.  

Ensuring that our policies are just and equitable when responding to risk is imperative, even if as individuals we struggle to interpret risk accurately.

Add new comment

Filtered HTML

  • Allowed HTML tags: <a href hreflang> <img src alt height width> <em> <strong> <span> <cite> <code> <ul type> <ol start type> <li> <dl> <dt> <dd>
  • Missing filter. All text is removed
  • Lines and paragraphs break automatically.
  • Web page addresses and email addresses turn into links automatically.
CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.
Image CAPTCHA
Enter the characters shown in the image.