Biostatistics and Epidemiology | Nerona, Anne Martha A.
Introduction and - Methods of tabulation, graphical
presentation, computation of averages, measures of variability Variables 2. Inferential Statistics - Refers to the methods involved in order to Mrs. Carmelita Atamosa make generalizations and conclusions At the end of the topic the students must be able to: about a target population based on results 1. Identify correctly Statistics and Biostatistics, from a sample, includes estimation of and its importance in the field of health. parameters and the testing hypotheses. 2. Determine different variables in the gathering of data that can affect or help in the study of a Biostatistics certain sample or population. • Statistical processes and methods applied to the collection, analysis, and interpretation of Statistics biological data and especially data relating to • The science that is making sense of human biology, health, and medicine. information and data around us. • A branch of applied statistics directed • Measure of a characteristic or attribute of a towards applications in the health sciences group of people or a sample population. and biology. • A branch of mathematics that involves in the • An innovative field that involves the design, collection, analysis, interpretation, presentation analysis and interpretation of data for studies and organization of data that will eventually in public health and medicine, to arrive at lead to the practical utilization and translation conclusions about disease and health risks by of data to meaningful and purposive actions or evaluating and applying mathematical and deeper and broader appreciation of realities. statistical formula to the factors that impact health. General Usage 1. In plural sense: it is used to refer to a set of Applications of Biostatistics data or a mass of observation. 1. Tool in decision making process: Example: a. Different people make decisions in Public Health: quantitative data needed different ways, some based on as basis for planning, monitoring and intuition, on past experience, or on evaluation of health services, which includes: existing information. Vital Statistics: data on vital events b. An information-based decision making like the number of deaths, births and process invariably needs the marriages. application of Biostatistics. Health Statistics: cover a wide variety 2. Development of new drugs. of numeric and information including 3. Investigation in the development of disease. morbidity (causes and frequency of illness). 4. Integral part of the foundation upon which the 2. In singular sense: it refers to the body of expertise of health administrators, planners methods or techniques for the organization and public health practitioners in general rests. and analysis of collected information. 5. Problem identification, assessment, allocation of funds and evaluation of programs Two Branches of Statistics PLANNING – MONITORING Characteristics in the Context of Singular Sense) – EVALUATION – DECISION
1. Descriptive Statistics Major Division of Statistics
- Refers to the different methods applied in 1. Mathematical Statistics – study and order to summarize and present data in the development of statistical theory and methods form in which will make them easier to in the abstract. analyze and interpret. 1 Biostatistics and Epidemiology | Nerona, Anne Martha A. 2. Applied Statistics – application of statistical • Examples: weight, age of gestation, attributes methods to solve real problems involving to certain issues, blood pressure, educational randomly generated data and the development attainment. of new statistical methodology motivated by • We assign variables to understand specific real problems. characteristics of a population we are studying. Basic Steps 1. Making Observations (gathering of data) Types of Data 2. Generating a Hypothesis (the underlying law Quantitative Data and order suggested by the data) 3. Deciding How to Test the Hypothesis (what • Data that can be measured (quantified) critical data is required) • Can be written down numerically 4. Experimenting (or observing) this leads to an Discreet inference that either rejects or affirms the • A count that can’t be made more precise, hypothesis typically involves integers (whole numbers) • Sets of data that record actual, physical things Data • Categories can be measured and ordered • Observations made on particular elements of according to amount a sample or population. • Examples: # of patients admitted, # of • Could be quantities or attribute of a sample or students in a class population that are either measured or Continuous observed through a process of collection for • Information that can be counted are the purpose of analysis. expressed with infinite degrees of precision Primary Data • Can be reduced to lower units (fractions and • Collected from the original source, first hand. decimals) • Original data collected specifically for the • May take any value, within a defined range purpose in mind. • Examples: height 160cm, weight 150.5 lbs) • Field researchers Qualitative Data Secondary Data • Descriptive data, difficult to measure or count • Collected from another purpose in mind. and cannot be written down numerically. • Desk researchers • Can be counted but mathematical operations could not be applied. Constant and Variables • Examples: severity of disease – light, Constant moderate, severe. • Is a phenomenon whose values remain the same from person to person, from time to Types of Variables time or from place to place. Dependent Variables • Has fixed value or definite value. • Outcome of interest which shed in response • Examples: the pull of gravity, # of months in to some intervention a year, minutes in an hour, speed of light. • Also called explanatory or predictor variables • Can be derived from the physical rather than Independent Variables the biological. • The result or outcome of the investigation Variable • Response variable • A phenomenon whose values or categories Intervening Variables cannot be predicted with certainty. • Brings an effect to the outcome of the study • An entity that can take on different values. which may work independently or in tandem • Information which changes in value from one with the independent variable. observation to the other. 2 Biostatistics and Epidemiology | Nerona, Anne Martha A. • Destroys the predictive ability of the • They don’t allow for relative degree of independent variable this it has to be difference between controlled. • Possess the property of magnitude but not • Types of Intervening Variables property of intervals o Moderators – have a contingent effect • Examples: guilty or not guilty, completely on the relationship between the agree, mostly agree, mostly disagree, independent and dependent variables. completely disagree, stage 1, stage 2, stage 3, o Mediators – describe how rather than stage 4 when the effect of the independent • Help answer the question: How is A bigger variable will occur on the dependent than B variable. • Peculiar characteristic is it can be ranked or ordered Types of Data as to Levels / Scale of Interval Scale Measurement • Exact distance between two categories can be In the process of distinguishing variables, the process determined but the zero is arbitrary of measurement is to be done. • This type of scale allows for the degree of Nominal Data difference between items, but not the ratio • Either measurable or ranked but simply between them. categorized or classified. • Possess the property of magnitude or intervals • Examples: name, address, student’s course) but not the property of rational zero. • Numerical values may be assigned but they • Examples: Temperature (arbitrariness of zero don’t hold any mathematical value. (Ex. male point is shown by the fact that) degrees = 1, female = 2) Centigrade does not mean absence of • Would help answer the question “Is A temperature at all; it is simply a reference different from B” point for purposes of measuring • Simplest nominal categories are existential • Would answer the question: By how many variables (Ex. with liver cancer – no liver units do A and B differ? cancer) Ratio Scale Nominal Scale • Similar to the interval scale but the zero point Two groups of nominal scale: is fixed and the ratio of two numbers can be 1. Nominal-Renaming meaningfully computed and interpreted. (most - occurs when each object in the set is measurements are in this category) assigned a different number • Possess all 3 properties: magnitude, interval, - example: Social Security Numbers and rational zero 2. Nominal-Categorical • Examples: weight (0 is always absence of wt, - occurs when objects are grouped into 100kgs is 2x 50 kgs). subgroups and each object within a • Would answer the question: How many times subgroup is given the same number. bigger than B is A? - the subgroup must be mutually exclusive, that is, an object may not belong to more NOTE: It is important to distinguish the type of than one category or subgroup. variable one is dealing with because it is one of the - example: based upon political party (Lakas major determinants of the type of statistical technique -1, Laban -2, Nationalista -3) that can be most appropriately applied to the data. - we cannot compute for the mean, SD Ordinal Scale • Are shown simply in order of magnitude since there is no standard of measurement of differences.