The Effect of an Intervention to Alter my Eating Behaviors Background

I have decided to alter my own habits surrounding consumption of fruit. The intervention I will use as an attempt to eat more fruit is by setting alarms in my phone for 11:00 a.m., and for 4:00 p.m. I believe this intervention will be useful to alter my habits because it will actively remind me to eat fruit at least once per day. The alarms are set at convenient times given my schedule. The baseline and intervention observations will be measured by two indicators; how often I wish to eat fruit, or crave fruit, during the day aside from the moment the alarm goes off (frequency); and, how often I follow through and do eat fruit at any point during the day (magnitude).

My goal for when I implement the intervention, is to increase my consumption of fruit by at least one (1) serving per day in order to improve overall health. Additionally, because it is thought that after 21 days an individual can form new habits, I hope that after 20 days I will feel close to forming a new habit and thus will continue to eat more fruit per day without the intervention method.

Methods:

The behavior I chose to experiment with is the amount and frequency of consuming healthy foods that I do. I will measure my consumption of servings of fruit because this is something very tangible to measure, easily accessible, and physically beneficial to the test subject and researcher, in this case me. Thus, experimenting with the magnitude and frequency to which I consume fruit can be operationalized and measured effectively over twenty days. The first ten of the twenty days will be spent measuring my Baseline (BL) results. The second set of ten days will be spent measuring behavior when an intervention is introduced to the environment.

The indicators I will use are (1) how often I think about eating fruit, and (2) how many servings of fruit I actually eat. The first indicator will measure frequency of the behavior. I expect the number of times I think about eating fruit (Indicator 1) to be higher than the times I actually do consume fruit (Indicator 2). Additionally, I expect there to be a significant amount of variation in the data for Indicator 1 because I might be inspired by my environment to think more often about fruit, such as seeing a friend eat fruit, or because I have one serving and then think about having more later. The second indictor will measure magnitude. My tracking the number of servings of fruit I eat per day, I can measure the magnitude of the impact that the intervention has on my behaviors. I chose to measure frequency and magnitude because I can see to what degree I follow through with actions.

The intervention I chose to implement to alter my baseline measurements is an alarm. I set an alarm for 11:00 a.m. and for 4:00 p.m. This was done to remind myself to eat fruit at some point during the day. I chose to set the alarms for 11:00 a.m. and for 4:00 p.m. because these times are around when I would normally want to eat or have a snack. Additionally, they are times when obtaining fruit would be easy because grocery stores would be open.

The only difficulty I faced when implementing this intervention plan was that if I was reminded to eat fruit because of my alarm but was not able to, or was not hungry and then decided to do so later, I would then forget. Therefore, the intervention was not one-hundred percent effective every day. This occurred on the twelfth day of the experiment, resulting in not eating any fruit. I also had not eaten any on the second and seventh day of the experiment. These two incidents can be explained because I had not been receiving the intervention treatment and therefore I was not frequently reminded to eat fruit and I forgot.

The intervention in the experiment did have positive effects as well as causing me to eat more fruit. I felt healthier and wanted to eat other healthy foods throughout the day. I did not want to drink soda or juice as much because I was getting natural sugar from the fruit. And, another minor result was that I did not feel bloated or sleepy after eating as much. Whereas when I eat less healthily, like eating pizza, I feel somewhat sleepy afterwards. Therefore, I think that the intervention to encourage myself to eat more fruit helped form other healthy habits and I feel like it will be easier to continue to eat fruit every day.

Findings:

Both graphs demonstrate significant change with the indicators when the baseline was challenged by the intervention. The intervention increased the behaviors frequency and magnitude to an average of 4.5 and 1.7 respectively from initially being an average of 2.9 for frequency, and 1.2 for magnitude. Due to this increase of 1.6 in frequency and 0.5 for magnitude, I would say that the intervention acted successfully through this experiment and that there was clinical, visible, significance from the interventions effects.

The most dramatic difference was between the frequency of which I thought about eating fruit. This is likely because it is much easier to think about doing something rather than following through with the action.

The physical improvement, the magnitude of the amount of fruit I actually ate, in the behavior was also visible. Aside from the qualitative aspects of my observation such as feeling healthier overall, I ate on average 0.5 more pieces of fruit per day when the intervention was in place.

Z-scores determine how close to the mean that an observation is and is measured by the Standard Deviation (SD). Therefore, when the mean is either higher or lower than the mean SD, there is clear statistical difference between them (Hanneman, Kposowa, A. J., & Riddle, 2012). Therefore, because the mean was 4.5 and the SDs are 0.62844 and 5.17156 for the first indicator, it has a statistically significant effect. Indicator 2 is also statistically significant. The mean is 1.7 and the SDs are -0.54 and 2.94.

Discussion:

The strengths of this research design were that it was very easy to control. This is because the experiment was kept to a simple, single system and there were not many external issues that may have deviated it from being completed. Because of its simplicity, the experiment makes it easy to define the causal relationship between the observed behaviors, thinking of eating fruit and then actually eating fruit, and the intervention which acted as the treatment for this experiment. The strength of how the intervention was assigned was that I could set the alarms for 11:00 a.m. and 4:00 p.m. on day eleven out of twenty and then have them set to repeat every day. This made the intervention easy to continue throughout the experiment because I did not have to worry about forgetting to implement the intervention. By taking out this aspect of human error, I controlled for potential biasness in the data. If the intervention had not been automatic, then some days during the intervention period would have been treated differently and I likely would have not eaten fruit. Thus, the before-and-after comparison between observation periods would have been skewed downwards and I would not have found such a dramatic, positive result.

The weakness in this experiment is in the data collection method. My data collection was done in my phone’s note-taking section. Every time I thought about eating fruit I would make a tally mark in the notepad, and the same went for every time I did eat. The problem with this is that I believe I might have left a few occurrences of thinking about fruit out of the notes, and therefore the data. This would be because I had been aware from my phone for an extended period of time, had thought about eating several times, and then was distracted with something else when getting back to my phone and therefore I did not record the frequency with complete accuracy.

My experiment demonstrates both internal and external validity. It demonstrates internal validity because I have a clear causal effect from the intervention. The average times I thought about or ate fruit increased during the intervention time. This also shows that the intervention was successful because it resulted in the desired effect. Additionally, my experiment demonstrates external validity and could be generally recognized for its causality. This is because the fact that the intervention worked to remind me to eat more fruit (as well as increasing my thoughts about it because I was more conscious of doing so), the intervention method (the alarms) could be used to alter other behaviors.

Conclusion:

Overall, this evaluation exercise was a beneficial experience. The intervention I implemented clearly had a positive and tangible effect on me throughout the second set of ten days. I believe that my method of intervention, the alarms, were an effective way to make changes in my life and could be applied to other’s as well. The benefits of this type of intervention was that I was conscious of the intervention happening which raised my overall awareness of the issue at hand, healthy eating, with the specific target being fruit. Because of this, I have formed the natural habit of eating more fruit, or fruit-based foods such as natural smoothies, fruit leather’, and fresh juices. The burden of this style of intervention was that I often felt guilty if I could not turn my attention to the experiment’s goals at the time of the alarms. However, this was minor and I do not believe it had an effect on the outcome of the experiment. If it did, however, I believe the outcome would be that I might eat more fruit than planned because I would not want to miss out. Therefore, overcompensation would be an issue. However, I do not believe this occurred in this experiment. I believe this type of research design is applicable for social work because one can very visibly tell the results of implementing an intervention into an individual’s life. It is easy to track data because the data collection would either be tallying up the total times someone engaged in an activity, or merely having a binary variable (yes/no, 0 or 1) for their action. An example of a binary variable a social worker may track would be, given this intervention, men with drug problems showered more frequently throughout the week’. This would be binary because the observer, or the subject, that is tracking the data would merely have to write yes’ or no’ that the subject showered on each day they are being observed.

Appendix:

Data on frequency: Data on magnitude:

References:

Hanneman, R., Kposowa, A. J., & Riddle, M. (2012). Basic statistics for social research. San Francisco, CA: Jossey-Bass.

Did you like this example?

1137

47