Presentation is loading. Please wait.

Presentation is loading. Please wait.

COL NAILA AZAM Biostatistics INTRODUCTION. LEARNING OBJECTIVES To understand the RELATIONSHIP OF BIO STATISTICS TO PUBLIC HEALTH To correlate collection.

Similar presentations


Presentation on theme: "COL NAILA AZAM Biostatistics INTRODUCTION. LEARNING OBJECTIVES To understand the RELATIONSHIP OF BIO STATISTICS TO PUBLIC HEALTH To correlate collection."— Presentation transcript:

1 COL NAILA AZAM Biostatistics INTRODUCTION

2

3 LEARNING OBJECTIVES To understand the RELATIONSHIP OF BIO STATISTICS TO PUBLIC HEALTH To correlate collection of data to basic requirement of VITAL STATISITCS To clarify THE TYPES OF DATA collected during research

4 What Is Public Health? “ Public Health is the science and art of preventing disease, prolonging life and promoting health through the organized efforts of society. ” (World Health Organization)

5 The Functions of Public Health Assessment: Identify problems related to the public’s health, and measure their extent Policy Setting: Prioritize problems, find possible solutions, set regulations to achieve change, and predict effect on the population Assurance: Provide services as determined by policy, and monitor compliance Evaluation is a theme that cuts across all these functions, i.e., how well are they performed?

6 What is Biostatistics? Statistics is the art and science of making decisions in the face of uncertainty Biostatistics is statistics as applied to the life and health sciences

7 The press frequently quotes scientific articles about:  Diet  The Environment  Medical care, etc. Effects are often small and vary greatly from person to person We need to be familiar with statistics to understand and evaluate conflicting claims

8 VITAL STATISTICS VITAL STATISTICS

9 VITAL STATISTICS A branch of statistics that deals with the changes and most basic events of human populations: e.g., natality (birth); mortality (death); morbidity (illness and disease); injuries; marriage... Vital statistics are indispensable in studying social and health trends, and making important legislative, commercial (marketing) and health decisions Statistics are gathered from census and registrars’ reports, physicians’ records, medical examiners’/mosque records, grave yards, and a variety of other health professionals

10 RATES: DENOMINATORS AND NUMERATORS Rates or ratios are used to measure most health problems They consist of numerators and denominators: a count of events divided by the number of possible events Numerators and denominators used in public health statistics are of three types:  Survival data (births, deaths, and a count of the population  Health and socioeconomic status data  Data based on health resources and utilization

11 THE DENOMINATORS The most important information on which activities in public health must be predicated is a count of the population to be served The decennial census is an important and the most widely used information as the denominator(defacto/dejure)  May include intercensal estimates, based on projects or sample surveys Other sources of information may also be used, depending on the phenomenon of interest; for example --  School enrollment records  Employer records of numbers of workers

12 THE DENOMINATOR  Airline carriers for numbers of passengers carried during a given time Numerator data stem from administrative registration and reporting procedures Most significant of activities relate to the vital events of birth, mortality (death), and morbidity (disease, illness, and injury

13 TYPES OF DATA Just as we must classify and organize information before we can retrieve and use it, We must classify data into the correct type before we can do any statistical analysis on them.

14 WHY? The data type will determine :  How data can be coded for analysis?  What kind of analysis can be performed?

15 CATEGORICAL DATA Nominal data: variables are divided into a number of named categories without any intrinsic order  Sex, marital status Ordinal data: variables divided into number of ordered categories  Level of knowledge, opinion on a statement NUMERICAL DATA( expressed in numbers): possible values take a distinct series of numbers Discrete continuous

16 NUMERICAL DATA EXAMINED THROUGH Frequency distribution Percentages, proportions, ratios, rates Figures Measures of central tendency Measures of dispersion

17 EXAMPLES 1. while checking accuracy of clinical diagnosis of malaria, data frequency distribution of 33 slides is,  Negative-19  P. falciparum-13  P.vivax-1 DATA ??  NOMINAL

18 2. 148 students were asked about attacks of palpitation on a scale:1/2=rare; 3/5= occasional; > 5 =frequent Never=47 Rare=71 Occasional =24 Frequent =6 DATA????? ORDINAL

19 FREQUENCY DISTRIBUTION OF NUMERICAL DATA (GROUPING OF DATA) Select groups for grouping the data Count the number of measurements in each group Add up and check the results GROUPING RULES:  Groups must not overlap  No gaps (continuity in groups)  Groups range from lowest to highest measurement; round numbers for lower values)  Equal width for copmarability

20 VARIABLES A variable is a characteristic of a person,object,or phenomenon that can take on any or different values  Age in years,months,weeks  Weight in Kg,pounds,stones,mg  Distance in m, km, walking minutes  Monthly income

21 Dependent, independent, confounding variables The variable used to describe or measure the problem under study  DEPENDENT VARIABLE The variables used to describe or measure the factors that are assumed to cause or at least to influence the problem  INDEPENDENT VARIABLE Variable related to both above variables  CONFOUNDING VARIABLE

22 EXAMPLE CAUSE EFFECT  (independent variable) - dependent variable OTHER FACTORS  Confounding variable

23 QUESTIONNAIRE 1. gender male------ female------- 2. postal code of your home address----- 3. how will you rate the quality of your mobile service provider?  Poor, fair, average, good, excellent 4. how many points have you accumulated for the bonus program? 5. how many SMS did you send yesterday? 6. how old are you? 7. what is your yearly income?  80000

24 NOMINAL DATA The answer to Q 1 will be either male or female, but before analyzing the data,we must code them by symbolically assigning a number to each possible answer. As 1 for female, 2 for male. Allowing limited calculations, no averages, only FREQUENCY COUNTED.

25 POSTAL CODE ? To find residential location, again postal codes are labeled with numbers No meaningful calculations allowed ; only simple counting.  So nominal data again

26 ORDINAL DATA The 3 rd question asks to rate the quality of service on a 5 point scale, the we can code the answers as 1 for poor and 5 for excellent(we can reverse the coding if we want) The coding is not arbitrary,i.e 1 for poor and 5 for excellent by compulsion  We must only follow an order, from high to low or low to high  Indicating rank of inherent quality(4 better than 3, 5 more than 4)

27 Still we cant say Difference between 1 & 2 is equal to difference between 3&4. Numbers indicate only show order(better /worse), not how much better Analysis of such data will again be restricted to frequency estimation only

28 Q NO. 4 introduces INTERVAL DATA A more familiar example would be that of temperature scales of CELSIUS AND FARENHEIT (WHERE THERE IS NO ANSOLUTE ZERO, ZERO POINT IS arbitrarily chosen and an object at zero is not without heat. [an object at 32 degree F is at zero degree C. An object at 20 degree is 5 degree less than one at 25 degree but we cant say that 20 degree is twice as warm as 10 degree.

29 INTERVAL DATA Interval data has no absolute ZERO point So we cant use comparisons as to twice as much or half as many with interval data.

30 Q NO 4, of BONUS POINTS The svc provider gives 2000 points to each user and then 100 points for each SMS. Users can claim various prizes on the basis of points collected. As absolute zero is missing; earning 10000 points is not twice as much as earning 5000 points  10000- 2000= 8000  5000-2000=3000

31 IN Q NO 5 &6 --RATIO DATA The answer to both questions is HOW MUCH GREATER OR LESSER. There is an absolute zero point so  zero SMS means no SMS

32 RATIO DATA Provide information detailed information on HOW MUCH GREATER OR LESSER  WE can USE COMPARISONS AS TO ‘TWICE AS MANY ‘ with confidence and surety.

33 DATA CONVERSION What about Q no 7???  With absolute zero / no income at all?????  Are the incomes to be treated as ratio data?????  Are these interval data ???????????

34 HOW ???? Respondents are asked to place themselves in one of the six categories from 1-6 with 1 as lowest and 6 as highest. If you are in 4 th cat and your friend in 2 nd cat, then your income is higher than him/her Again we cannot say how much? So we are collecting  ORDINAL DATA

35 ANY QUESTIONS????? FEELING READY FOR RESEARCH PROJECT??????


Download ppt "COL NAILA AZAM Biostatistics INTRODUCTION. LEARNING OBJECTIVES To understand the RELATIONSHIP OF BIO STATISTICS TO PUBLIC HEALTH To correlate collection."

Similar presentations


Ads by Google