Limitations and Delimitations. This study contains several forms of sample bias. First, the study utilizes responses from APIC members, as it is comprised mainly of infection control professionals. However, the members are a sample of convenience, and are not necessarily representative of all infection control professionals. Second, although the study included information on participants’ facility, the survey collected data at an individual level. Therefore, it is possible that multiple individuals from the same institution completed the survey, causing an oversampling that could affect the analysis. The latter point is also a limitation in the study design. In order to ensure no identifiers, i.e., personal or institutional, were captured, the study design focused on an anonymous, individual level survey in lieu of a facility-based survey. The potential for social desirability bias exists in this study as well. A majority of participants indicated that they were aware of CDC HICPAC guidelines, which could be an over- exaggeration as infection control practitioners may be embarrassed to admit they had no knowledge of the CDC HICAPC guidelines. Further, the survey contained a link to both the CDC HICPAC guidelines and the norovirus prevention toolkit, as a means to promote and improve their knowledge. This may have contributed to the large proportion of participants that were aware of the CDC HICPAC guidelines. However, if that is the case, then the awareness of the norovirus prevention toolkit may be considerably less than reported. The survey response rate for this study was low (5.4%). The survey distribution coincided with the APIC annual conference, which may have affected the response rate. Additionally, survey fatigue could have an effect on the response rate since APIC members are utilized quite frequently for surveys. However, response rates from other surveys that utilize similar populations, i.e., APIC members, have comparable response rates (Allegranzi et al., 2014; Xxxx et al., 2014; Xxxxxxxxx et al., 2012), suggesting this is a typical response rate. Improvement of the survey tool could better elicit responses and analysis of results. For example, skip logic directed respondents that selected “Other” as a facility to question 2 to complete the number of beds their institution contained. Some facilities, such as public health departments, should have been exempt from completing this question. To alleviate confusion in responses, a “Not Applicable” response should have been in...
Limitations and Delimitations. The study took place at a small private, not for profit, Christian, liberal arts, four-year University in Southern California. First-time first-year college students participated in the study. Findings and recommendations based on the survey results are specific to this institution, which limits generalizability and implications of the results. Data collected for this study took place during the summer/fall 2016 and fall 2019 semester. The study will only report findings and conclusions on students attending these semesters and participating in the study. The Financial Education Proficiency developed for the study included questions adapted directly from Xxxxxxx’x (2008) Jump$tart college student survey. The Financial Self-Efficacy Scale questions (Xxxx, 2011) allows for the evaluation of a respondent’s awareness of their behavior and consequences associated with making specific financial decisions. Although these instruments have been validated with other samples, they were not created for college students, and this may impact the reliability and validity of the data. For each instrument, choice of wording and order were maintained and only slightly altered for a college student audience. In some instances, readability, word choice, and flow may have affected survey responses. Regardless of these limitations, the study sheds light on the much-needed topic, the importance of the relationship between college student financial literacy, self-efficacy, and financial planning.
Limitations and Delimitations. 70 CHAPTER 4: RESULTS ...................................................................................................... 71
Limitations and Delimitations. 82 Conclusions ............................................................................................................... 83 REFERENCES ...................................................................................................................... 84 APPENDICES ....................................................................................................................... 93 Appendix A: Teacher and Curriculum Coach -Part Survey ...................................... 93 Appendix B: School Leader Part Survey................................................................... 95 Appendix C: School Leader Questionnaire ............................................................... 97 Appendix D: Teacher Questionnaire ......................................................................... 98 Appendix E: National Institute of Health-Dr. Xxxxxxx Xxxxxxx ................................ 99 Appendix F: National Institute of Health-Xxxx Xxxxx 101 Table 1. 1. Demographics of Participants........................................................................ 45 Table 2. 1. Evolving Themes ........................................................................................... 45 Table 3. 1. My Relationships with the School Leaders are Positive ............................... 46
Limitations and Delimitations. A noticeable limitation in this project is the lack of a “true” dataset. A “true” dataset is the complete dataset without any missing observations. This is impossible to acquire since missing data may be due to systematic issues, lack of response, miscoding, errors in imputation and a variety of other factors. Consequently, the mechanism of missing data needs to be investigated. Complete-case analysis assumes that the data is either MCAR or MAR. This assumption is robust to small amounts of missing data (<5%); however, large proportions of missing data are unlikely to be MCAR. This project assumes that the missing data mechanism was MAR. It is also possible that the missing data mechanism was NMAR. An assumption of Case study 1 is that MPR accurately reflects patient adherence to their statin therapy. There are limitations with this assumption. MPR does not directly measure patient consumption of their statin therapy; instead, it provides an indirect estimate of adherence based on pharmacy refill data.76 Other forms of adherence measurements are available which were not used in Case study 1. For example, Proportion of Days Covered (PDC) reflects the percentage of days the medication was available to the patient.64 PDC is calculated as the total days the complete medication regimen was available divided by the total number of days evaluated capped at 1.0.64 MPR can also be truncated or allowed to exceed a cap of 1.0. In Case study 1, MPR was truncated at 1.0. It is unclear whether using the PDC would impact whether a patient was adherent or not. Xxxx, et al. reported that differences between MPR and PDC were negligible and provided similar answers in terms of categorizing patients as adherent or non- adherent.64 Another limitation of this project is the small sample size of the liraglutide group relative to the exenatide group in Case study 2.18 The small sample is potentially sensitive to missing data which can result in inaccurate parameter estimates due to large uncertainties or variances. A larger sample size would mitigate this issue; however, there was not possible with the current design. In observational studies, unmeasured variables can be potential confounders despite controlling for all measurable variables. Propensity score matching may be considered in this situation, however it is highly sensitive to unmeasured confounders.77 It is not an absolute answer in the absence of a randomized controlled trial. Future investigation using propensity sco...
Limitations and Delimitations. As with any modeling exercise, the necessary simplification of a complex reality implies limitations that must be considered in the application of the results. Since there is almost no surveillance data that would specify the etiology of diarrhea during a CHE we had to estimate the disease burden. We did not take into account the effects of herd immunity of rotavirus vaccination, which may lower disease burden. The model also did not consider possible vaccine side effects, including intussusception that might become apparent with large-scale implementation of the vaccine. However, they were not observed in the clinical trials of Rotarix.74,75 Given the probable high burden of rotavirus disease in CHE, even if these were to occur they are very unlikely to significantly alter the cost-effectiveness of the vaccine. The range of vaccine effectiveness was estimated from El Salvador and may not be a true estimate of the coverage that can be achieved in Somalia during a SIA. Given the high rates of diarrhea, the vaccine effectiveness may be on the lower end of the range. It has been observed with oral polio vaccine, another live vaccine, among populations with high burden of diarrhea, malnutrition, and other medical conditions such as TB and HIV,32 the vaccine effectiveness is lower. A decrease in vaccine effectiveness would increase the CEA.
Limitations and Delimitations. Due to the ongoing COVID-19 pandemic, we were not able to recruit, nor perform interviews in person. Additionally, during the pandemic it was difficult to get respondents during the study recruitment period. Due to the limitations on recruiting in person and low response rate through emailing organizations, social media was the primary recruitment method. A meta-analysis on the use of social media for recruiting participants into studies found evidence that it can be the best recruitment method for observational studies and for individuals within specific groups and with specific conditions (Xxxxxxxxx-Xxxxxx & Xxxxxxxxx, 2016). We also found using Facebook and Instagram ads was an effective recruitment strategy for our specific population and age group (Xxxxxxxxx-Xxxxxx & Xxxxxxxxx, 2016). There may be limitations to the results since the population sampled from mainly consisted of users of Facebook and Instagram. The people selected were known to have internet use, since they were using some form of social media, there may be skewed telehealth connectivity results. Due to the retrospective nature of the data gather, there may be potential recall bias. There is always a possibility that participants willing to participate are different than the general population, but this is hard for researchers to mitigate. The risk level of COVID-19 may differ among patients, however, from what was gathered from the respondents, they all had the same basic COVID-19 safety protocols implemented within their care locations. We were limited to using Zoom for our interviews because of the pandemic, which may have altered the interactions between interviewer and interviewee, but also may have made the interviews more accessible for participants. We did not accurately access ethnicity since we paired it with race and respondents mainly chose their race without choosing the ethnicity, so only race can be compared but not ethnicities within the results and conclusion. The study did not have the ability or resources to create the online consent, screener, or interviews in languages other than English, so this non-English speaking population is omitted from the study and language proficiency is not a consideration for this study in the satisfaction of perinatal telehealth services. The study aimed to have equal number of rural and urban in-depth interview participants, but by using the Atlanta Regional Commission definition of what classifies a county as urban and rural, our inte...
Limitations and Delimitations. Potential weaknesses identified in this project’s design include lack of generalizability of findings, opportunities for missing data, inability to address all barriers to care, and uncertain long term impact of the intervention. This project was carried out in Georgia’s WCHD among HIV+ women enrolled in the RWP overdue for cervical cancer screening. Since the intervention’s benefit will be to this particular group, generalizable knowledge will not be developed. Chart review and CAREWare were the sources of outcome data and may not contain a complete record of the Pap tests performed. Pap reports from outside providers may not have been requested, or if requested, may not have been received. In addition, all available Pap results may not have been entered into CAREWare. Not all of the perceived barriers to this population’s care could be addressed, thus some women still had obstacles to completing the screening test. Finally, it is uncertain whether the intervention will have a lasting effect on the patient’s completion of future Pap screenings. Several factors narrowed the scope of the project. These factors relate to the selected group of participants and the timeframe for the intervention. The target population was comprised of HIV+ women enrolled in Georgia’s WCHD RWP overdue for cervical cancer screening. The inclusion and exclusion criteria used were as stated in the HAB performance measure. All women ≥ 18 years old or who reported a history of sexual activity and had had a medical visit with a provider with prescribing privileges at least once in the measurement year were included in the project. Women were excluded if they were < 18 years old and denied a history of sexual activity or if a hysterectomy had been performed for non- dysplasia/nonmalignant indications. The project was conducted from September 1, 2012 through December 31, 2012.
Limitations and Delimitations. Since the study was conducted within a pre-selected population that was already part of a cohort study, the sample used is not completely random and therefore decreases the generalizability of our findings. This study is also limited by the self-reported nature of cross-sectional surveys that can introduce information bias. Certain questions in the survey asked about past events, which can lead to recall bias. In addition, questions about hygiene and sanitary conditions are subject to bias since participants may not always feel comfortable sharing personal information. Furthermore, due to logistical and financial constraints, water quality data was only collected at one point in time and not longitudinally, which could have provided more insight into water quality variability, especially if data had been collected in more than one season.
Limitations and Delimitations. Some bias may have occurred during the data collection process, the facilitator was a contract worker for WFP. While every effort was made to create rapport and ensure confidentiality, it is possible that certain interviewees did not feel comfortable discussing dissatisfaction with someone from outside their community who was associated with the program. Ideally the interviews would have been conducted in secluded settings but the first priority was to conduct discussions in convenient locations for the women. Discussions were often interrupted by local community members, which may have compromised the participant’s feelings of anonymity. Additionally, there are some issues related to data quality: the focus group participants could not always be identified during transcription. This is a result of the note taker aiding in the facilitation process and was unable to note the speaker. Furthermore, the facilitator often phrased questions in a leading manner, which may have biased answers. This qualitative study had a very narrow focus and only investigated the barriers, knowledge, attitudes, and practices of mother beneficiaries. Mothers who are not enrolled in the stunting prevention program were not included which means these results are not externally valid.
