Question 2 - John
In the article by Riebl et. al, there are several limitations of the study that are discussed. What stands out to you as the most significant limitation and how could the researchers redesign the study to overcome these limitations?
posted by Unknown @ 3:10 PM
25 comments
25 Comments:
The most significant limitation seems to be the cross-sectional nature of the study. Longitudinal studies can be expensive and difficult to gain approval for though. I think the cross-sectional nature and the researcher’s acknowledgment of this limitation serves to validate and support the need for a longitudinal study. This study demonstrated great value in better understanding this relationship between different influences and an individual’s beverage choices. If we can figure out what the most prominent causal influences are during different periods in an individual’s life, we can better harness that influence to ensure good messages are being sent and reinforced (even at the level of the future choices these adolescents make in building their home beverage environment with their own kids). It would be really interesting to see if changing the availability of SSBs especially at home would have an influence down the line. Starting now might trigger rebellious behavior, but if the child never became familiar with SSBs in the first place, would they really miss it or think it was such a social norm?
I have to agree with Gina the study design chosen is the most significant limitation. A longitudinal study would have provided more data. However, the current study did address the importance of the issue and the need for more studies on the subject. Perhaps in the future a longitudinal study could be conducted. I think when the study is conducted a larger sample size should also be considered. Gina mentioned also looking at the availability of SSBs, which is a great idea. This goes back to modeling children’s food and beverage choices growing up. If the child doesn’t see it around or see their parents drinking it will the child be more likely to consume it themselves? Most likely, they won’t be. I think it would also be interesting to look at the health warning labels mentioned in the supplemental article and see how the labels would affect adolescents decisions.
As both Moriah and Gina mentioned, a major limitation of the study is the nature of the design being a cross-sectional study. In addition, I also thought their second limitation of there being no superior method of diet evaluation, which also goes hand-in-hand with the study design. As we know, 24 hour recalls are not always the most accurate means of gathering data about a person’s diet. Ideally we would have had a longitudinal study performed where participants track their intake as they go. However, very seldom do you see such study designs in place.
I think the greatest limitation is the lack of diversity in the population. 93% of the adolescent participants were white and therefore the study’s outcomes may vary in a different, more diverse population. Since the population was not diverse it would be hard to generalize the results to other populations and therefore additional studies need to be conducted with different diverse populations. I believe food preferences vary greatly among different demographics and cultures. For example, in a population with a poor socioeconomic status food preferences might be more influenced by what the family can afford vs a population of wealthy individuals might be more influenced by taste or peers.
I agree with Amanda, Moriah, and Gina that a major limitation of the study is the cross- sectional design. However, longitudinal studies are not very realistic because of their expense as well as the time commitment for these studies. I think focusing on diversifying the population would be a much easier way to improve the limitations of this study as opposed to conducting a longitudinal study because it would take much less time and money.
I agree with everything the others have said regarding the weaknesses of this study: it's cross-sectional design, lack of dietary analysis, the availability of SSBs. Moriah brings up a great point that while looking at adolescent food and beverage choices is important, it can all relate back to the choices that were available to them as children and how their guardians monitored their food choices and health education. This could be looked at as a form of primary prevention, if you will, for ensuring that kids continue consuming healthy things as they gain freedom in food choices during adolescence.
Some limitations to the study that I noticed that I was surprised the authors did not note is the lack of diversity of the sample. The subjects were predominately white, from middle to upper class households, and had highly educated parents. This limits the results to only applying to a minor portion of the population. This place of high privilege puts a whole new spin on the meaning of the results and ignores some of the stressors and influences that may influence peers of other socioeconomic statuses.
It sounds as though we are all in agreement that the cross-sectional design of the research study was one of the most limiting factors. As Gina mentioned a longitudinal study perhaps would have provided better understanding of the relationship between different influences and an individual’s beverage choices. Carly also brings up a great point that there was a clear lack of diversity within this study that may certainly have affected this study's outcomes.
Great discussion so far!
A limitation discussed in the article that I found significant to the results was that the population for the study came from a single region, therefore lacks diversity. Having samples from different regions with different SES as well as race/ethnicity is important for future implementation of interventions catered to the general adolescent population. As Carly mentioned, people of different ethnic backgrounds may have grown up with certain dietary habits or food preferences. With that said, a study focused on a population from a single regions maybe difficult to standardize to people of different cultural backgrounds.
I agree with Justine, as families of high SES may have different education backgrounds, and may be more aware of the dangers of SSBs. They may have the knowledge on what foods/beverages to purchase that are lower in sugar, and substitutions for SSBs for their children, compared to low SES families that may not have the education or resources to do so.
I agree with Carly that the lack of diversity in the study population is the most significant limitation to me for the same reasons Carly had mentioned. It would probably be inaccurate to say most adolescents make the same choices and are influenced by the same factors in the same way as the study population of upper class, educated, white families. Nikki brought up a great point that the education about SSBs can greatly differ between the two populations, which can strongly influence one's decisions. This limitation definitely warrants for a more diverse sample population in a repeat study.
I feel that all the limitations of the study are significant. However, I feel that the lack of diversity may be most significant. As Carly and Justine mentioned, 93% of the population was classified as white, which lacks minority populations that have been shown to drink more SSBs than other populations (which was mentioned in the supplemental article). Additionally, the lack of diversity extends to show the lack of overweight and obese individuals in the sample. As mentioned in the article by Riebl et al. (2016), the sample distribution of overweight and obese individuals is well below the state average, which further indicates that this study is not representative of the state or the entire adolescent population. I feel it is important to note this particular limitation because overweight and obesity can correlate with other lifestyle factors, such as increased screen time, decreased physical activity, and an overall poor diet quality (http://www.nrcresearchpress.com/doi/abs/10.1139/apnm-2013-0551#.VsPKPZMrLBI).
I agree with Gina and Moriah's point regarding the benefits of using a longitudinal study. Due to resources, it is usually not feasible, especially when a study like this would need to be conducted with a much larger, more diverse sample population. I wholeheartedly agree with Vivian regarding the replication of this study for more accurate results.
Limitations of this study include the cross-sectional view of the adolescent diet within 4 distinct day samples. However, due to the variety of age distribution a correlation may be evaluated among shifts in beverage consumption throughout the adolescent time span sixth grade to senior year within the scope of the study. More pertinent is the lack of direct correlation between parent behavior or parenting strategies and child beverage habits as described as a study focus. Parents were described as completing health histories, questionnaires, and food frequencies upon consent, yet not sample number was provided as to how many of the parents chose to consent and participate with their child. Also, little effort was expended to limit contributing variables such as physical activity level, parent education level, or perceived health status in analysis.
Researchers may tweak the study design to include more contributing variables for elimination. Also, specifying the sample population would provide more standing in generalizing the data.
I see the largest limitation to this study as being in one isolated geographical location. Because this study only had a sample from the Blackburg, VA area, it is difficult to generalize the studies findings to the entire adolescent population in the USA. SSB consumption greatly varies across different regions of the country. For example, when completing an undergraduate internship in Nome, Alaksa, the prevalence of SSB consumption was much greater than I had ever experienced in the lower 48.
In response to Carly's comment, the lack of diversity of the sample with having 93% of the sample population as white adolescence, the limitation of isolating the study to one geographical location truly stands out. There are found trends between different ethnicity's and intake. It really is a cause and effect limitation. Because the study only surveyed adolescence from the Blackburg, VA area, the sample reflected the demographics of the people in that city and surrounding area.
Noel, I thought you made a great point about the geographical location being a big limitation. SSB is similar to food consumption and as we know that can vary from state to state. We also see rates of obesity differ state to state which shows location plays a part in these findings. Although many research studies are focused around one area, due to convenience, it would be important to further study different geographical locations to make this study more credible and make the findings more valid.
Like Noel, Carly you also mentioned a limitation that I thought was very important to think about. America is a very diverse Country and having nearly all the population of participants be white certainly lacks diversity in the study and could affect the findings of the study. Food preferences most defiantly vary between different cultures. There are also people that come from different Countries where SSB is not a norm. Thinking about diversity, the design of the study, and the location there are many limitations in the study and all are very important to consider.
I believe that the cross-sectional design is the most notable limitation. The impossibility of determining causal relationships is an area that weakens the study’s findings. The cross-sectional approach only captures a snapshot of the overarching factors that could be influencing the SSB intake habits of the adolescents studied. SSB intake habits could be largely influenced by the adolescents’ historical relationship with such drinks. Their childhood exposure to SSB and what they associate with such beverages can influence their adolescent intake regardless of their adolescent intentions/attitudes or parental influence. For example, I grew up associating SSB with holidays, birthdays, and eating out. They were not a part of my daily life early on, and thus were not a regular part of my diet in adolescence despite peer influence. A longitudinal study would be preferred for studying this topic, but this may have cost, time, and compliance limitations. Aside from changing the study design, the researchers could not do very much to overcome this limitation. The lack of causal relationship determination is inherent in cross-sectional studies.
As several of you have mentioned, I believe that the study’s largest limitation was its lack of diversity. The study’s adolescent participants were 93% white and parent participants were 86% female, 97% white, and 88% married, also, 80% of participants were of higher SES. Additionally, the study only had a sampling of 7% obese or overweight, roughly half of the sampling area’s population (15% overweight, 12% obese). Since food and beverage choices are known to vary across demographics and regions, the study would benefit from a larger and more diverse sampling. The researchers could overcome these limitations by utilizing stratified sampling instead of convenience sampling.
This comment has been removed by the author.
I would have to agree with Noel that the largest limitation is that the sample population was from one geographical location and therefore not being diverse. The United States has become a melting pot of different cultures and this sample did not accurately represent that. I think it would be more realistic if they were able to gather data from a more diverse group. It would be interesting to even compare different locations/populations as well.
It appears that almost everyone agrees that the largest limitation of this study is absence of diversity. I agree with this, as well. This study provides results for the population of adults and adolescents in Blacksburg, Virginia. The participants sampled were predominately white, and of normal weight. I also think it is very important to point out that 97% of parents were highly educated. Studies have shown that those with higher educations and socioeconomic status have better health outcomes. In the future, a similar study should be conducted with a more diverse population.
I think the cross-sectional design is the most significant limitation to the study. From the single, snapshot nature of the study and its participants, there is limited diversity and invalidated dietary recall data as discussed. By replicating the study with different geographic sectors and/or completing a more long-term study over time many of these limitations are likely to be addressed. The methodology of the study does not appear to be a major limitation at this point, but it is more a test of reliability and validity to determine if findings are able to be generalized to larger populations.
There will always be limitations to studies which is unfortunate but unavoidable. The goal is to try to minimize the limitations and their effects. I think Nikki and Viv both made great points about the population from the study. After talking with interns who have worked in a low income setting, I think that SSBs are more prevalent in that population. That could skew study results. I think that selecting a diverse population including gender, ethnicity, SES, and age would help minimize inaccuracies. I also agree with Gina that a longitudinal study would have been a more ideal study design, but a cross-sectional study is more realistic. I think that this limitation is part of research since it is not always possible to conduct a longitudinal study.
The most significant limitation to the study would be the use of the cross-sectional design. To have a more valid and credible study, a longitudinal study should be utilized and with a larger sample size with an increase of diversity among the population.
I agree with the fact that was brought to our attention by Justine and Nikki. The chosen population, families of high SES, may have different education backgrounds, and may be more aware of the dangers of SSBs. They may have the knowledge on what foods/beverages to purchase that are lower in sugar, and substitutions for SSBs for their children, compared to low SES families that may not have the education or resources to do so.
-Kelly Berg
I agree with Carly that the biggest limitation in this study is the lack of diversity. I am currently at a rotation site where the patient population is significantly different than what I am used to working with at different sites. When I started there, I quickly learned that their diet recalls, specifically with their SSB consumption varied greatly from other populations that I have worked with. Patients at my current site report in their diet recall that they are drinking 1-2 pitchers of kool-aid or sweet tea a day or 6-10 non-diet sodas a day. This kind of variation could have played a significantly role in swaying the results of this study.
Like Kristen and Rachel (and a few others) mentioned, I think the cross sectional study design is a major limitation. I feel like to gather real insight about the topic at hand researchers would have to repeat the study several times with the same participants to glean trend data that causation hypothesis could be founded upon.
I can see where the geographic limitation people are coming from also. Only narrowing in on one physical location can hinder the generalization of results of the study. I think the best way to overcome these challenges would be to repeat the studies multiple times (longitudinal data collection) in several geographically diverse locations over time.
Noel made a great point about the limitation of the study’s limited geographical location. SSB consumption varies from state to state and within each state. The limited sample location and lack of diversity compound the limitations of the study, making the results hard to generalize to the whole Virginia adolescent population let alone the entire US adolescent population.
One of the greatest limitations of this study is its sample population. It is widely known that socioeconomic status is highly related to food choices. The researchers were appropriate in measuring the obesity levels of the sample in Blacksburg Virginia, but the samples obesity average was 5-8% less than the national average, making it difficult to generalize or apply its results to the general population. This is a common limitation of research studies, as widening the sample population can be unattainable for various reasons. However, this studies results provided valuable information to the literature and should be repeated to reflect a more diverse population and to address the numerous other limitations.
Post a Comment
<< Home