Research Methods in Dental Studies

Introduction to Dental Research

• The scientific method is based on reproducibility, feasibility, and reliability.
• A systematic process involves deductive reasoning, including formulating a research question, predicting a hypothesis, designing a plan, testing the hypothesis, and evaluating results.
• Choosing the research approach depends on the nature of the problem and phenomenon being studied, with the deductive approach being prevalent in dental research.
• Method refers to tools, experiments, techniques, and tests used to collect and analyze data, while methodology is broader and encompasses various methods to investigate research hypotheses.
• Methodology involves the correct application of methods, rationale behind method selection, and criteria for choosing a method.
• Variables are measurable factors involved in answering the research question, including independent variables (cause or risk factor) and dependent variables (effect or outcome).
• An association (correlation) exists between variables when changes in one variable coincide with shifts in another, and it can be positive or negative, proportional or nonproportional.
• A causal association occurs when changes in the independent variable cause a difference in the dependent variable.
• Hypotheses are statements in response to research questions, guiding the research process and being validated through experimental testing.
• Participants/samples are drawn from a population for research studies, and sampling involves choosing portions of the population to observe and study.
• The sampling method can be probable (random) or non-probable, and it's important to differentiate between random sampling and random assignment/allocation.
• Hypotheses play a guiding role in research, and participants/samples are drawn from a population, with sampling being a crucial process in research studies.

Research Design Cycle Key Points

• Investigate exposure to a specific agent or therapeutic/diagnostic intervention
• Comparative intervention or control groups are considered in the research design
• Intended outcomes such as curing, prevention, reducing pain, etc., are clarified
• Factors of time, like age of patients or duration of intervention, are sometimes considered
• Example research question: "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)?"
• The hypothesis is formulated based on the research question
• Hypothesis provides a framework for answering the research question through statistical tests
• Observation of existing literature forms the basis for developing the hypothesis
• Factors to consider for formulating hypotheses include statement of relations, verifiability, and relevance
• The number of hypotheses may vary based on the nature of the study and its specific aims
• The null hypothesis (H0) is derived from the research question and serves as a comparison point
• The proposal serves as a guide in conducting, monitoring, and evaluating the research.

Research Design Cycle Overview

• The design cycle of research consists of four steps: investigation, designing a plan, performing the research, and research evaluation.
• The investigation phase involves creative thinking, critical reading, and formulating a hypothesis through analysis, formulating research questions, and decision-making.
• Analysis of the collected information is crucial in the investigation phase to examine adequateness, completeness, and knowledge gaps.
• The formulation of the research question is essential for interpreting and converting the research topic into a question that guides the research.
• Different types of research questions are related to the type of research, such as exploratory, descriptive, observational analytical, and interventional research.
• A good research question should be original, pertinent to current scientific knowledge, precise, coherent, concise, and feasible, without ethical issues.
• Formulating the research question can be done through considering the PICO framework, which involves patient/population, intervention, comparison, and outcomes.
• The PICO framework includes defining the patients/populations, intervention, comparison, and outcomes for the research question.
• The intervention component of the PICO framework involves defining the treatment, exposure, causative agent, or diagnostic test.
• The PICO framework serves as a guide for formulating research questions by considering the specific aspects of the research.
• The research question should be clearly defined to guide the research process and outline the necessary steps to answer the question posed.
• Following the design cycle is essential to ensure proper results and avoid issues during the research process.

Determination and Definition of Research Questions

• A good research question should be clear, focused, succinct, complex, and argumentative.
• It initiates and paces the research project, guiding the design, methodology, and hypothesis.
• Research questions are vital for both qualitative and quantitative research, providing a way to navigate the writing and research process.
• There are two types of research questions: qualitative and quantitative, and they guide the type of research and data collection.
• The initial step in research design involves identifying a gap and creating a focused research question.
• It is important to select an interesting topic, conduct preliminary research, and consider the audience when developing research questions.
• Asking questions and assessing the effectiveness of the research question are crucial steps in the process.
• The research question should be specific, centered on the main topic, and tackle a difficult problem that requires in-depth analysis.
• A research question must be formulated in a way that allows the research to solve part of the problem identified.
• The research question should be able to find an answer using primary and secondary data sources and be possible to resolve within given time and other constraints.
• Different research designs and methods can be used to address various research questions, but they must all be pertinent to the study's objectives.
• Writers can avoid "all-about" papers by asking straightforward and specific research questions that help them focus on their research and support a specific thesis.

Ethics in Animal Studies and Manuscript Writing

• Russell and Burch proposed three principles for animal research: replacement, reduction, and refinement.
• Replacement involves developing alternative methods to replace animal experiments.
• Reduction involves testing the minimum number of animals required for valid results.
• Refinement focuses on providing good laboratory conditions to reduce animal suffering.
• Additional considerations include the necessity of animal experiments, choice of animal species, and ethical proposal writing.
• Approximately 115 million animals are used for dental research annually.
• COPE is a non-profit organization investigating ethical issues in scholarly publishing.
• Authorship in scientific articles is guided by the International Committee of Medical Journal Editors (ICMJE).
• Authors must fulfill specific roles to be included as authors in a manuscript.
• Journals may require authors to define their roles and contributions in an "author statement file."
• Fraud in research includes data fabrication, manipulation, and omission.
• Undeclared conflicts of interest can affect the judgment of reviewers or editorial boards.

Bias and Confounding in Dental Studies

• Bias influences study estimates, leading to overestimation or underestimation of true population values, stemming from study design or execution.
• Subject-related bias and measurement errors are sources of bias in dental studies, impacting study outcomes.
• Selection or allocation bias and response bias can affect study validity due to the chosen participants and differences between responders and non-responders.
• Recall bias, withdrawal bias, and follow-up bias are biases from patients, potentially distorting study conclusions.
• Assessment bias and interviewer bias are biases from researchers, influencing how results are rated and interview conduct.
• Clinical trials are less prone to biases due to randomization, blinding, and intention-to-treat analysis, while observational studies are more susceptible.
• Confounders are variables associated with both the outcome and exposure, leading to misrepresented or masked relationships in studies.
• Age, gender, and time are common confounding factors in dental studies.
• Randomization in clinical trials aims to create treatment arms similar in characteristics, minimizing confounding effects, whereas observational studies use methods like matching during sample selection to reduce known confounders.
• Observational studies require complex analyses, like multiple regression, to account for multiple influencing factors and determine if effect sizes might be influenced by bias and confounding.
• Dental journals mainly feature phase III clinical trials that assess treatment effectiveness in substantial patient numbers, with some phase II trials focusing on preliminary safety and efficacy assessments.
• Cross-over trials in dental studies compare treatments within the same patient, minimizing needed patient numbers compared to parallel trials, and provide identical treatment groups to minimize biases.

Understanding Study Design and Sample Selection

• Informed consent is a legal requirement for human studies, especially in clinical trials, and ethical considerations are crucial.
• Experimental studies involve active interventions, while observational studies observe without intervening.
• Clinical trials test treatment effectiveness, while observational studies identify risk factors and causes of diseases.
• Observational studies involve random sampling, while clinical trials invite participants meeting specific criteria and allocate them using randomization.
• Random sampling requires a representative frame from the population of interest, while random allocation in clinical trials defines a target population based on inclusion and exclusion criteria.
• Sample size is crucial for study design, and statistical techniques help determine the appropriate sample size to detect clinically important effects.
• The significance level determines the threshold for accepting or rejecting the presence of an effect and is often set at 5%.
• Power gauges the likelihood of detecting a specified difference if it genuinely exists and is typically set around 80% or 90%.
• Cross-sectional studies capture information from a sample at one point in time, lacking the longitudinal aspect of cohort studies and retrospective data collection of case-control studies.
• Cohort studies are less prone to bias compared to case-control studies, but they can take many years and are less commonly used for rare diseases.
• Case-control studies are prone to selection bias in choosing control groups, impacting study validity.
• The objective in all studies is to ensure similarity between groups with and without the disease, except for exposure, to minimize bias and confounding factors.

Understanding Data Types and Data Analysis

• Variables can be dependent or independent, with the former relying on another variable's value and the latter not being influenced by another variable.
• Data represents observed values of a variable and can be categorized into nominal, ordinal, interval, and ratio data types.
• Nominal data includes categorical variables without inherent order or ranking, while ordinal data has an inherent rank or order.
• Interval data has ordered values with meaningful differences, and ratio data includes properties of interval data along with an absolute zero point.
• Descriptive statistics are used to provide simple summaries about the sample and form the basis of quantitative data analysis.
• Data collection techniques include interviews, questionnaires and surveys, and examination of documents and records.
• Interviews can be formal, semi-structured, or informal, and are mainly qualitative in nature.
• Questionnaires and surveys allow for analysis using quantitative methods and are easier to analyze than qualitative techniques.
• Examination of existing data, such as databases and records, is an inexpensive way to gather information but may be incomplete.
• Data can be visually presented using graphical methods like bar diagrams, pie diagrams, line charts, and tabular methods such as frequency and cumulative frequency tables.
• Summary measures provide information about sample data, falling into two main categories: measures of location and measures of spread.
• Measures of location include mean, median, and mode, while measures of spread include range, variance, and standard deviation, providing information about the variability of the data.