Lecture 2 PDF - Research Methods in Dentistry

DENT 4407 Research Methods in Dentistry Design Cycle of Research Dr. Gülşilay SAYAR, DDS,PhD Associate Professor Bahçeşehir University School of Dental Medicine Department of Orthodontics [email protected] References Research Methods In Dentistry Fahimeh Tabatabaei Lobat Tayebi Des...

DENT 4407 Research Methods in Dentistry Design Cycle of Research Dr. Gülşilay SAYAR, DDS,PhD Associate Professor Bahçeşehir University School of Dental Medicine Department of Orthodontics [email protected] References Research Methods In Dentistry Fahimeh Tabatabaei Lobat Tayebi Design Cycle of Research What Is the Design Cycle? Detailed observation of a problem, creative thinking, and critical reading about it generate a design cycle. When you come across a math problem, you read the question carefully several times (understanding the question is very important). Next step, several solutions come to your mind to solve that problem. You usually try to choose the one that will get you to your destination faster. After that, you will solve the problem based on that solution and finally check it to ensure your answer. “Problem-solving” is a conscious, logical, thought-provoking, and purposeful process that helps you to find practical solutions to problems through it. Design Cycle of Research What Is the Design Cycle? The selection of a research topic leads to the launch of a design cycle that is based on the problem-solving method. It is a series of steps that a researcher takes to start working on a project. The design cycle consists of four steps: (1)İnvestigation (formulating the hypothesis), (2)designing a plan (writing the proposal), (3) performing the research (hypothesis testing), and (4) research evaluation (data analysis and interpretation) It is essential to stick to these steps because skipping one or doing them out of order will lead to improper results. Design Cycle of Research Investigation (Formulating the Hypothesis) When we choose a topic for new research, creative thinking and critical reading of the existing literature and the current situation can give us insight into the prognosis of our study. Design Cycle of Research Investigation (Formulating the Hypothesis) Now is the time to investigate and formulate the hypothesis, which includes four basic steps: (1)analysis of the collected information, (2)formulating the research question, (3) decision-making about the different ways to reach the answer, and (4) formulating the hypothesis. The research importance in following a design cycle focuses on the “investigation phase,” This investigation determines the overall direction of the research and should be very accurate; otherwise, the researcher will face many issues during the next steps of research. Design Cycle of Research Design Cycle of Research Analysis of the Collected Information After we have access to lots of information about the background of the problem and the methods already used for dealing with the problem. It’s time to examine the adequateness and completeness of the information collected from the previous steps. This step is crucial in the “investigation” phase. Design Cycle of Research Analysis of the Collected Information By analyzing the collected data, you can clarify the knowledge gaps, define the constraints that may affect your results while they cannot be changed, examine whether the problem can be broken down into smaller sub-problems, and consider potential solutions for each sub-problem. The scientific evidence gathered by literature review, the researcher’s experience and skills, the research center’s priorities, may influence the “data analysis”. Design Cycle of Research Design Cycle of Research Design Cycle of Research Formulating the Research Question Formulation of a research question means interpreting and converting the research topic into a question that can be answered. Research questions develop and explain the identified problem. The research topic will be formulated to a specific question that clearly defines the general objective. If you present the research topic without specifying the research question, it would not be apparent “how new knowledge will be generated.” Design Cycle of Research Formulating the Research Question The research question plays a guiding role because it will serve as a basis for formulating and identifying the hypotheses. Developing the research topic into a question makes it easier to outline the steps necessary to answer the question posed. It helps the researcher to understand more clearly what he/she is looking for. By framing the question, the problem is almost half solved. It is also helpful during proposal writing for answering the questions raised. Design Cycle of Research The type of question is wholly related to the type of research In exploratory research (case report), question is about what happened (e.g., what are the adverse effects of tongue piercing? In the descriptive research (cross-sectional study), you will ask about the characteristics of the problem (What are the prevalence and risk indicators of periimplantitis? Design Cycle of Research In the observational analytical study (case-control, cohort study), you will ask about the relation of two or more variables together (Is there any association between orofacial pain and depression? The question in the interventional (experimental) research is about the effect of one variable on another one (Does preemptive analgesia with intravenous acetaminophen affect postoperative pain after third molar surgery? Design Cycle of Research A good research question has these features: • The question should be original (novel) and engaging. The originality of a research question means it has not yet been investigated from a certain angle you want to research. • The question should be pertinent to current scientific knowledge (educational relevance), to clinical and health policy (clinical relevance), or to future research (research relevance). • The research question must be precise, coherent, and concise with a definite focus. Design Cycle of Research A good research question has these features: • A good research question is feasible. Feasibility means that the number of subjects, the number of variables, and the technical skill for answering the question are adequate; it needs reasonable time and money . • There should not be ethical issues in the research question or in answering it, resulting in the rejection by the institutional review board (IRB). It is therefore advisable to find out about the necessary authorizations before writing the proposal. Design Cycle of Research However, the best way to formulate the research question is to consider four parts of the PICO framework. Patient/Population, Intervention, Comparison, Outcomes Design Cycle of Research P: the patients, populations, problem, subjects, or samples. For example, for evaluating the relationship between periodontitis and amalgam, you must define what group of society your question is about. People with a specific disease (patients with diabetes)? People in a particular service setting (dentists, dentist assistants)? Or a specific group of people (amalgam factory workers)? Design Cycle of Research I: the intervention (treatment, exposure, causative agent, diagnostic test). Here you should describe if you want to investigate exposure to a specific agent (capsules containing mercury and amalgamator) or a therapeutic or diagnostic intervention (amalgam filling/amalgam replacement). Design Cycle of Research C: the comparison intervention (no treatment, no exposure, alternative treatment) or the negative and positive control groups. You may also consider the gold standard as a control group for comparing groups. If the study group were patients with an amalgam filling, the control group could be persons without filling or having a composite filling. Design Cycle of Research O: the outcome or intended results. You should clarify what your intended outcome is. Is it curing, prevention, reducing death, reducing pain, recovery, the incidence of side effects, reducing the duration of hospitalization, change in the satisfaction rate, or improving the quality of life? Which outcome is to be affected? For example, the incidence or increase of periodontitis could be the outcome. Design Cycle of Research T: sometimes, you should also consider the factor of time like the age of patients, duration of intervention, or the time outcome will be reviewed. Design Cycle of Research In endodontic retreatment of patients with peri-radicular lesions (P), does photodynamic therapy (I) have more effect than the 810 nm diode laser (C) in reducing the number of the microbial flora of the root canal (O)? Design Cycle of Research Design Cycle of Research Decision-Making About the Different Solutions for the Research Question Design Cycle of Research Design Cycle of Research Formulating the Hypothesis In the last step of the investigation phase, the research question must be translated into the hypothesis, and the researcher must imagine the anticipated answers to the Question. Design Cycle of Research Design Cycle of Research The objective of a hypothesis is to provide a framework for answering the research question through statistical tests. The value of scientific work depends heavily on the formulated hypothesis. If a researcher has enough knowledge, he/she can make predictions by formulating hypotheses. It should be emphasized that hypotheses should not be lightly conjectured but must reflect the investigator’s essential knowledge, imagination, and experience. Design Cycle of Research The observation of “what already exists” and the literature analysis is the basis for developing the hypothesis. Its formulation must be precise so that the main objective of the Project meets the identified problem. The hypothesis statement makes it possible to present the reasons or rationale for the experiment. Design Cycle of Research The factors to be considered for formulating the hypotheses are: The statement of relations: In most hypotheses, two main concepts are considered: causes or factors (independent variables) that affect dependent variables. Design Cycle of Research If you specify the direction of the difference between groups in your expected outcome by terms such as “less than,” “greater than,” “positive,” and “negative,” you have a one-sided hypothesis. In a two-sided hypothesis, which is preferable, there is no specification about the expected direction of the difference. In this case, the meaning of the relation is indicated by terms such as “different from.” Design Cycle of Research Verifiability: The essence of a hypothesis is that it can be verified. It contains observable variables which can be measured and analyzed. Increasing the number of variables would increase the time required for data collection and complicate the study and the data analysis. Design Cycle of Research Relevance: The hypothesis must be relevant to the phenomenon under study. Design Cycle of Research In scientific research, depending on the nature of the study and its specific aims, the number of hypotheses may be one or more. We can have a central hypothesis and secondary hypotheses. The number of hypotheses is not imperative, but their coherence and coordination are essential. Secondary hypotheses must be articulated around the central one and call each other in a logic imposed by the research problem. Design Cycle of Research When you formulate a research question based on the PICO elements, you can quickly acquire your hypothesis by changing it into a statement; by adding “not,” you can then turn it into a null hypothesis (H0) During the research process, the researcher’s hypothetical answers to the research questions (H0) are accepted or rejected (or the state of uncertainty remains unresolved, especially when the sample size is too small to ensure reliability). Design Cycle of Research We can refuse the null hypothesis when the study results are statistically significant (a statistically significant difference between the variables). Answering a research question can lead to the emergence of a new question, constructing another null hypothesis, and continuity of the research. Design Cycle of Research Designing a Plan (Writing the Proposal) The proposal will serve as a guide in conducting, monitoring, and evaluating the research. If you want to conduct a survey, research questions, main objective and specific aims, relevance, methods of data collection and their justification, study population, sampling, list of resources, anticipated answers, and tools envisaged to evaluate data should be explained in the proposal. Design Cycle of Research Design Cycle of Research Performing the Research (Hypothesis Testing) Once the research problem has been sufficiently developed, the research plan has been written, and the necessary funding has been received, it is time to perform the research. Design Cycle of Research The data collection process requires preliminary steps such as authorizations to conduct research, training of team members, managing the budget, strategies for data collection, management of potential problems Design Cycle of Research An experienced researcher, based on the literature review and the prepared proposal, knows well what techniques and tools he/she should use to measure the variables and collect the data. The types of tools can be: • Interviews, questionnaires • Measuring devices such as calibrated probes • Films for X-ray images; radiographs • Cellular/microbiological techniques Techniques for analyzing tissue samples Design Cycle of Research Organizations like the International Organization for Standardization (ISO) or American Dental Association (ADA) have defined several standardized methods and procedures for data collection, which routinely are used in dental research Standardization is a process by which methodological differences between studies can be minimized. Design Cycle of Research Evaluation (Data Interpretation) Once the research plan has been performed, you should evaluate your data based on the objective and success criteria that you have defined. By performing the research, you will have access to some raw data. To evaluate the collected data and find the answer, you must: (1)put these data in specific tables based on specific goals, (2)check whether the differences are significant or not by performing various statistical tests, and (3) finally interpret and discuss them. Therefore, the evaluation step consists of three phases: processing, analyzing, and interpreting data. Design Cycle of Research Design Cycle of Research Normal distribution means following a unimodal, symmetrical, or bell-shaped distribution on both sides around the mean You can check the normal distribution of data by statistical tests such as Kolmogorov-Smirnov and Shapiro-Wilk tests or graphically using percentile-percentile (P-P) or quantile-quantile (Q-Q) plots Design Cycle of Research If the distribution is non-normal, the data should be analyzed by nonparametric or distribution-free tests such as Mann-Whitney, Wilcoxon, Kruskal-Wallis, or Friedman. Design Cycle of Research Thank you...