Bad Design Yields Poor Results
To study the impact and effectiveness of a treatment upon a disease or medical problem, researchers must follow a defined approach to get the most accurate results. If a researcher takes an incomplete approach to the design in his or her research, the results will be inaccurate and can even be biased. This is the very problem with the vast majority of the anti-spanking research. Let’s take a look at an example to illustrate.
Example: If you are researching the effectiveness of an antibiotic against ear infections in children, here are the steps you would take.
- Problem: Ear infection is the problem to be studied. This seems obvious, but if not defined then the results will be very fuzzy.
- Tendency: Take a look at each child’s tendency to get ear infections based upon his or her unique characteristics. Some children are more prone to infection due to their family history, (parents or siblings had numerous infections and/or need ear tubes), allergies, or immunity problems. Children in daycare, compared to those who are kept at home, are more likely to have ear infections due to exposure to other children with “colds.” These factors will affect the success of the antibiotic.
- Dose & Delivery: Define the dose of the antibiotic and how it is given: by mouth, by injection, or by infusion.
- Measure of Result: At the end of the study, you would then examine the ears of each child in the study to judge how effective the antibiotic was in treating the infection. A less accurate measure would be to interview the patients to see how their ear feels after treatment, without actually looking at the ear.
- Comparison: Compare the effectiveness of this antibiotic to other antibiotics used to treat ear infections. This will determine whether the results are unique to the antibiotic being studied.
Now, let’s apply this same design to a study seeking to answer the question: How effective is spanking in correcting a specific misbehavior?
- Problem: Define the particular problem behavior being studied, rather than looking at general child behavior.
- Tendency: Determine how non-compliant (disobedient) each child is at the start of the study. Those who are more compliant will respond more quickly, with fewer spankings, and with a greater chance of success. The children who are more disobedient at the start of the study are more likely to be more disobedient by comparison at the end of the study, even if the spanking was judged to be effective. A child’s tendency to obey is in part determined by that child’s general temperament. For example: compliant vs. rebellious, patient vs. impatient, impulsive vs. thoughtful. Without this measure of tendency before and after the spanking, it could appear that those who are spanked more often (more rebellious at the beginning) become more rebellious after spanking when compared to those with more compliant temperaments that required fewer spankings.
- Dose & Delivery: Define the spanking technique and when it is used.
- When to spank
- How many swats
- Where to swat (buttocks only or also legs)
- Use it with or without explanation (reasoning)
- Use in public or only in private
- Use alone or along with other measures, such as time-out
- Only with certain ages or with any age child
- Measure of Results: Establish how the results will be gathered: from direct observation or from interview results of the child or parent? Are the results prospective longitudinal (data from direct observation as the study is conducted) or retrospective (data from interviews which depend upon the memory of the participants about events of the past).
- Comparison: The results of spanking must be compared with other measures (time out, privilege removal, grounding) to be sure that the results are unique to the use of spanking.
Current Research on Spanking
Spanking research that most closely follow the standards previously outlined have shown spanking by parents to have neutral to positive effects upon children. Almost all of the research claiming negative effects of spanking upon children have not followed these standards and do not hold up to rigorous evaluation.
Problem: The problem behavior to be measured is rarely identified, but when it is defined the results have been positive for spanking.
Tendency: Each child’s tendency to be aggressive or disobedient is not measured in any of the condemning studies. Therefore, child aggressiveness appears to increase in proportion to the frequency of spanking.
Dose & Delivery: This is the most glaring deficit of the anti-spanking studies. How, when and where spanking is administered is never specified. The parents are never taught how to use the measure. Its use is left to each parent’s whim and therefore is not uniform within the study. This approach allows for inappropriate usage to be included in the study group, thereby negatively affecting the results.
Measure of Results: The highest quality spanking studies, which are longitudinal, revealed the most positive results with spanking. The retrospective studies are vulnerable to interviewer bias and poor recall.
Comparison: Although spanking is rarely compared with other measures in anti-spanking studies, when it is, the negative findings against of spanking disappear.