Statistical psychology

To become proficient in statistical reasoning, you need to be able to interpret data from surveys. This includes preparing frequency tables and graphs, as well as histograms.

Imagine you are an intern with a community clinic. Your organization conducted a survey regarding patient satisfaction. Your supervisor has given you the original survey and the survey results and asked you to prepare a summary including the following:

Frequency table
Histogram
A frequency bar graph
A summary describing the trends and/or patterns observed in the data

Full Answer Section

        Hypothetical Survey Results (n=50 patients): Imagine the survey collected 50 responses, distributed as follows:
  • Rating 1 (Very Dissatisfied): 5 responses
  • Rating 2 (Dissatisfied): 7 responses
  • Rating 3 (Neutral): 10 responses
  • Rating 4 (Satisfied): 18 responses
  • Rating 5 (Very Satisfied): 10 responses

 

1. Frequency Table

  A frequency table organizes the raw survey data, showing how often each response occurs.
Satisfaction Rating Frequency (Number of Patients) Relative Frequency (Percentage)
1 (Very Dissatisfied) 5 (5/50) * 100% = 10%
2 (Dissatisfied) 7 (7/50) * 100% = 14%
3 (Neutral) 10 (10/50) * 100% = 20%
4 (Satisfied) 18 (18/50) * 100% = 36%
5 (Very Satisfied) 10 (10/50) * 100% = 20%
Total 50 100%

 

2. Histogram

  A histogram is used for visualizing the distribution of continuous or ordinal data. In this case, since our satisfaction ratings are ordinal (ordered categories), a histogram is appropriate to show the distribution across the satisfaction scale. Description of a Hypothetical Histogram: If we were to create a histogram based on the above frequency table, it would have:
  • X-axis (Horizontal): Labeled "Patient Satisfaction Rating," with numerical values 1, 2, 3, 4, and 5 marking the center of each bar.
  • Y-axis (Vertical): Labeled "Frequency" or "Number of Patients," ranging from 0 up to at least 18 (the highest frequency).
  • Bars: There would be five vertical bars, each representing one satisfaction rating. The height of each bar would correspond to its frequency.
    • The bar for Rating 1 would reach a height of 5.
    • The bar for Rating 2 would reach a height of 7.
    • The bar for Rating 3 would reach a height of 10.
    • The bar for Rating 4 would reach a height of 18.
    • The bar for Rating 5 would reach a height of 10.
  • Appearance: The bars would typically touch each other, indicating the continuous nature of the scale (even if discrete points are used). The shape of the histogram would show a slight positive skew, with the peak at "Satisfied" (Rating 4) and then a decline.

 

3. Frequency Bar Graph

  A frequency bar graph is very similar to a histogram for ordinal or nominal data, with the key difference being that the bars typically do not touch, emphasizing distinct categories. Description of a Hypothetical Frequency Bar Graph: If we were to create a frequency bar graph based on the above frequency table, it would have:
  • X-axis (Horizontal): Labeled "Patient Satisfaction Level," with categorical labels like "Very Dissatisfied," "Dissatisfied," "Neutral," "Satisfied," and "Very Satisfied."
  • Y-axis (Vertical): Labeled "Number of Patients" or "Frequency," ranging from 0 up to at least 18.
  • Bars: Five vertical bars, one for each satisfaction level. The height of each bar would correspond to its frequency.
    • A bar for "Very Dissatisfied" reaching a height of 5.
    • A bar for "Dissatisfied" reaching a height of 7.
    • A bar for "Neutral" reaching a height of 10.
    • A bar for "Satisfied" reaching a height of 18.
    • A bar for "Very Satisfied" reaching a height of 10.
  • Appearance: The bars would typically have a small gap between them. This graph would visually represent the same distribution as the histogram but is often preferred for discrete categories for clarity.

 

4. Summary Describing Trends and/or Patterns Observed in the Data

  Based on the hypothetical patient satisfaction survey data: The overall trend indicates that a majority of patients reported being satisfied or very satisfied with their experience at the clinic. Specifically, 36% of patients were "Satisfied" (Rating 4), and an additional 20% were "Very Satisfied" (Rating 5). Combined, this means 56% of patients had a positive experience. This is a strong indicator of general patient approval and suggests that the clinic is meeting the expectations of more than half of its patients. However, the data also reveals significant areas for potential improvement:
  • A notable 20% of patients reported feeling "Neutral" (Rating 3). These patients represent an opportunity for the clinic to convert lukewarm experiences into positive ones. Understanding the specific reasons for their neutrality could unlock pathways to higher overall satisfaction.
  • Furthermore, a combined 24% of patients reported being "Dissatisfied" (14% at Rating 2) or "Very Dissatisfied" (10% at Rating 1). This quarter of the patient population signifies a critical area that requires immediate attention. These negative experiences, though a minority, can have a disproportionate impact on the clinic's reputation, potentially leading to patient attrition and negative word-of-mouth.
Key Patterns Observed:
  • Bimodal Tendency (with a primary peak): While the highest frequency is at "Satisfied" (Rating 4), there's a smaller, but significant, cluster of patients at the "Neutral" and "Dissatisfied" ends, indicating that experiences are not uniformly positive.
  • Room for Growth: The significant percentage of "Neutral" and "Dissatisfied" responses suggests that while core services might be acceptable to many, there are likely specific aspects of the patient journey (e.g., wait times, communication, specific staff interactions, facility comfort) that are leading to less than optimal experiences for a substantial portion of patients.

Sample Answer

        As an intern with the community clinic, I understand the importance of interpreting survey data to inform our patient satisfaction initiatives. While I do not have access to the original survey and its specific results, I have generated hypothetical data for a common patient satisfaction question to demonstrate the required statistical summaries. Hypothetical Survey Question: "On a scale of 1 to 5, how satisfied were you with your overall experience at the clinic today?"
  • 1 = Very Dissatisfied
  • 2 = Dissatisfied
  • 3 = Neutral
  • 4 = Satisfied
  • 5 = Very Satisfied