1. Adding Points:
* Method: A fixed number of points is added to everyone's score. For example, adding 5 points to everyone's score.
* Advantages: Simple to implement. Raises all scores, improving morale.
* Disadvantages: Doesn't address the shape of the distribution. Doesn't account for differences in student performance. Inflates scores artificially, making it harder to compare across different exams or classes.
2. Percentage-Based Curve:
* Method: Scores are adjusted to match a target distribution. For example, aiming for a mean of 75% or a standard deviation of 10. This often involves rescaling scores based on z-scores or percentiles.
* Advantages: Allows for a more controlled distribution. Can be fairer than simple point addition.
* Disadvantages: Can arbitrarily shift grades. Students might get a lower grade than their raw score, even if they performed well compared to their peers. Requires some statistical knowledge to implement correctly.
3. Normal Distribution Curve:
* Method: Scores are adjusted to fit a normal distribution (bell curve). This often involves calculating z-scores and then transforming them back to a desired mean and standard deviation.
* Advantages: Creates a standardized distribution.
* Disadvantages: Assumes that the scores should naturally follow a normal distribution, which might not be realistic for all exams. Can penalize high-performing students if the distribution is already relatively normal or skewed towards higher scores.
4. Linear Transformation:
* Method: This involves adjusting the scores using a linear equation to map the original range of scores to a new range. For example, transforming a range of 40-100 to 60-100.
* Advantages: Simple and transparent. Guarantees a minimum passing score.
* Disadvantages: Doesn't consider the distribution of scores. May not accurately reflect relative performance.
5. Raw Score Percentile Conversion:
* Method: Each student's raw score is converted to a percentile ranking relative to the class. This ranking is then mapped to a letter grade based on a pre-defined distribution (e.g., top 10% get A's, next 20% get B's, etc.).
* Advantages: Directly reflects the student's relative performance within the class. Easy to understand.
* Disadvantages: Can lead to a harsh grading scale if the class performs exceptionally well or poorly.
Before Curving:
* Consider the exam itself: Was it too difficult? Were the questions fair and well-written? Addressing these issues before curving is often better than trying to fix a poorly designed exam with a curve.
* Transparency: Communicate your curving method clearly to the students *before* the exam or as soon as possible after. This builds trust and avoids misunderstandings.
* Justify your choice: Be prepared to explain your choice of curving method and why it's appropriate for the specific exam.
In short: The "best" curving method depends heavily on context. Avoid methods that arbitrarily inflate grades without considering the actual performance distribution. Prioritize transparency and fairness, and be prepared to justify your decision. Often, the best solution is *no curve at all* if the exam was well-designed and accurately assesses student learning.