Categories: Help me Write Essay

Statistical Error: Horn/Halo Effect

Statistical Error: Horn/Halo Effect

Definition: The tendency to allow one’s judgment of another person to be unduly influenced by an unfavorable (horns) or favorable (halo) first impression.

  1. Provide a Human Resource decision that has to be made as an example to the horn/halo effect.
  2. Errors typically occur because the data used to make the decision is flawed in some way. What flawed data could lead to the error for this decision?
  3. Think about what data could be used instead (to avoid the error: horn/halo effect?
  4. What parameter or statistic will you use to represent the dataset?
  5. How could this help avoid the error?

Statistical Error: Central tendency

Definition: The tendency of managers to give all employees average ratings. This reduces the variance in the data to the point where it becomes useless.

  1. Provide a Human Resource decision that has to be made as an example to the central tendency effect.
  2. Errors typically occur because the data used to make the decision is flawed in some way. What flawed data could lead to the error for this decision?
  3. Think about what data could be used instead (to avoid central tendency)?
  4. What parameter or statistic will you use to represent the dataset?
  5. How could this help avoid the error?

Criteria:

  • A decision is named. An actual decision must be named in the example.
  • The decision is relevant to HR. The decision should be relevant to one of the functions of HR (recruiting, selection, performance management, learning & development, compensation, safety, laws & regulations, etc.) or the overall HR strategy.
  • The flawed data would lead to the error. The ‘flawed’ data describe should reasonably lead to the error covered this week.
  • The alternate data would minimize the error. The alternate data (data to be used instead) is distinct from the flawed data and directly addresses the error.
  • The parameter or statistic is appropriate to the data. A parameter or statistic is what will be used to represent the full dataset and is based on the distribution of the data. Examples include weighted scores, means, or modes. The parameter chosen should reasonably represent the data.
  • Explanation of avoiding error. The explanation of how using different data could help avoid the error should be clear and accurate