Impact of hands-on training in STATA for data management and data analysis: How much knowledge gained among health-care professionals?

INTRODUCTION

Data management and statistical analysis are integral parts of any research. Data management is a process of collecting, cleaning, organizing, and maintaining data in a systematic and meaningful way for accurate analysis.^[1] It is very critical to have a quality data to provide best analysis which means data quality problems such as manual data entry errors, missing data, and equivocal data have to be resolved before analysis. There are many tools to identify errors and missing observations. One of the tools to identify this is STATA. According to StataCorp (2016), Stata is “a complete, integrated statistical software package that provides everything for data management, data analysis, and graphics.”^[2] Once, the data management is over, data analysis is performed.

For data analysis, researcher must have some basic knowledge of statistics. Statistics is a branch of science that helps to organize, describe, classify, and analyze the data and interpret the outcomes.^[3] Statistics is essential in every field ranging from commerce to healthcare, psychology to social sciences. Biostatistics is the use of statistical tools in health care and biological sciences. In medical education, biostatistics is an integral part of the curriculum. An innovative and computer-application-based learning is always aspirational and orients the learner in applying the tools toward relevant research and developing evidence-based interventions. It helps health-care professionals in accurate analysis and interpretation of results. Lack of biostatistical knowledge can lead to unethical research, and misinterpretation of results leading to a negative impact on evidence-based medical practice. There are many statistical software for statistical computing such as R (www.r-project.org), SAS (www.sas.com), SPSS (www.spss.com), and Stata (www.stata.com). Stata is both command line and point-and-click menus software. Stata is the most practical and effective software for data management and analysis in health-care research. Therefore, there is a need to introduce a training course on Stata for data management and data analysis for the health-care professionals and medical students.

We conducted a study to find out if providing a short-course training on STATA to healthcare eye professionals will achieve the objective of improving their competency to use computer-based statistical software packages for data analysis in addition to an increase in theoretical knowledge of data management and biostatistics. We also assessed effectiveness of the training using some key performance indicators such as facilitator’s knowledge, participant’s self-efficacy, and attitude toward training. We requested the learners’ feedback on challenges identified and suggestions to improve the training program.

MATERIALS AND METHODS

We organized a 10-day short course on data management and data analysis using STATA at the Indian Institute of Public Health, Hyderabad (IIPHH) under Operational Research Capacity Building project funded by Seva Foundation, USA. Eight participants from four eye care institutes (Nepal-3, India-1) received the training in STATA. It was a heterogenous group, comprising ophthalmologists, optometrists, outreach, and program managers.

The facilitators presented a brief overview of the topic at the beginning of each session. Care was taken to make the sessions engaging and memorable with frequent interactions through questions, exercises, and discussions using real-world examples. Using Google forms a pre and post-test evaluation was done every day. The questionnaire covered several areas of research methodology from data management to statistics. STATA software test assessment was done each day before beginning the session to assess pre-existing knowledge of the participants on the topics to be presented. The post-test questionnaire helped to evaluate the improvement in knowledge and skill gained through the course on that particular day.

On the last day of the training, the participants were asked to fill out the feedback survey form. The responses evaluated the facilitator’s knowledge, participants’ self-efficacy and attitude post-training, challenges encountered, and scope for improvement. No sampling method was followed. All participants included in the study were anonymized. Data were recorded in MS Excel and statistical analysis was carried out using STATA.

Categorical data were described in frequency and percentage whereas continuous data were described in median and interquartile range. Wilcoxon signed-rank test was applied to find out the significant improvement in the knowledge regarding the STATA software. Institutional ethical committee clearance was taken by the IIPHH. P < 0.05 was considered statistically significant.

RESULTS

Eight participants took part in the short course on STATA training. All participants held a graduate or post-graduate degree and worked in eye care institutions with an average work experience of 12.5 years (SD ± 6.0). The mean age of the participant is 37.9 years (SD ± 4.1 years). The female to male ratio is 1:3. Almost two-thirds of the participants had prior knowledge in managing data and applying basic statistical tools using other software such as SPSS, Epi-Info, and MS Excel.

Before providing training in STATA, pre-test evaluation was carried out every day. At the end of each day, a post-test evaluation was conducted. The 10 days cumulative median score for pre- and post-test was 43 (IOR = 39–47.5) and 72.5 (IOR = 69–77), respectively, and was statistically significant (P = 0.012). Male participants, younger age groups, experience, and prior knowledge were identified as major determinants to improve the knowledge and performance in usage of STATA [Table 1].

Facilitator evaluation

All participants completely agreed about the technical support provided by facilitators. About 87.5% of the participants agreed that the presentation given by the facilitators was clear, and valuable with various real life applications whereas 25% of the participants were neutral regarding the consistency among lectures, discussions, and exercises. The major reasons are basic knowledge of statistics and prior knowledge regarding systems/technology [Table 3].

Table 3: Feedback survey to evaluate the facilitator’s knowledge.

Facilitators evaluation	Agree (%)	Neutral (%)
Technical support was easily available when or if it is needed	100	0
There was a good balance between lecture, discussion and exercises	75	25
Various methods were used to present concepts or ideas	87.5	12.5
The course material was of high quality	75	25
Course faculty presented information clearly	87.5	12.5

About 62% of participants said that 80–100% of the training is linked to a real-world application followed by 25% who said 60–80% and 13% who said 40–60% [Figure 1]. After the training course, the major challenges for participants are too many competing priorities, followed by no time to apply the skills and no opportunity to use the skills [Figure 2]. Overall, 87.5% are likely and more likely to share this training with your colleagues. The participants also provided some suggestions to improve the course. These suggestions are: increase the duration time for this training, have basics knowledge on statistics, need a refresh course after some period and after that an advance level course on the same.

Close

Figure 1:

Proportion of training devoted to real-world application.

Export to PPT

Close

Figure 2:

Different challenges after the course training.

Export to PPT

DISCUSSION

Our 10 days hands on training workshop in data management and data analysis using STATA showed that there was a statistically significant improvement in knowledge of STATA software based on pre-test and post-test. A focused workshop on introduction to statistical methods and SPSS hands-on training to enhance analytical skills among research professionals held at Vijayapura by Tellur et al. showed improvement where the median score of pre-test was 12.5 and post-test was 14.0.^[4] In another study, Prabhu et al. showed an increase in mean scores from 3.42 (pre-test) to 10.53 (post-test) and Sari found an average pre-test and post-test score was 58.37 and 84.85, respectively,^[5,6] whereas in our study there was a substantial increase from 43.0 (pre-test) to 72.5(post-test).

We also assessed the facilitator’s knowledge, participants’ self-efficacy, and attitude at the end of the training. Majority of the (87.5%) of the participants agreed that adequate learning happened through assignments, quizzes, and knowledge checks. They also agreed that the facilitators had adequate knowledge of the course and presented the information clearly with various concepts and ideas. Similar results were found in a study by Prabhu et al. where the majority of the participants rated the sessions as “good.”^[5] We observed that an advanced training in usage of STATA is highly beneficial to people interested/involved in research to gain autonomy in data management and analysis.

In the present study, participants reported significant barriers to impending research and hence utilization of the STATA tool such as too many competing priorities in your work, no time to apply the skill and no opportunity to use the skill. Similar barriers to research had been reported in a study done by Giri et al. where lack of time was the major obstacle followed by inadequate financial support.

CONCLUSION

The evaluation of the training confirms that hands on training in the use of STATA software do have a significant impact on knowledge, self-efficacy, and attitude of the learner. Since Stata is very user friendly, we recommend that researchers and medical students should be offered a comprehensive training course in STATA.