How to Analyse Data
- 1. Analyzing Research Data AMIT SHARMA ASSOCIATE PROFESSOR DEPT. OF PHARMACY PRACTICE I.S.F. COLLEGE OF PHARMACY GHAL KALAN, FEROZPUR G.T. ROAD MOGA, 142001, PUNJAB
- 2. Analyzing Research Data • Data analysis can be divided into two: Descriptive and analytical. • In descriptive statistics, the main objective is to summaries the variables concerned one by one. • In analytical statistics the aim is to describe the relationship between two variables or more
- 3. Descriptive Statistics • The first thing to do before we analyze the data is to describe the variables. • When we are studying hypertension in one population, we should first describe the population. • We can describe the distribution of age, gender, race, education and income level. (Demographic Profile of Population)
- 4. Descriptive Statistics • We should include precision of the estimates, such as the standard deviation, standard error or confidence interval. • The estimates used to describe a variable depend on the level of its measurement. • For categorical measures, frequency or count with percentage we should use
- 5. Descriptive Statistics • E.g. distribution of gender, which is a nominal measurement, we could describe the number of male and its percentage over total number. • Usually in a variable with two categories data like this, describing one category is enough because the other will complement the value.
- 6. Descriptive Statistics • For numerical measures, it will depend on the Normality distribution • If it is a normal distribution, we could use mean and its dispersion values such as standard deviation. • When data are not normally distributed, we should not use mean and its product because mean is affected by extreme values.
- 7. Descriptive Statistics • In such situation, median is a better alternative. • We should not use standard deviation as it is derived from mean. • We could use minimum-maximum value, range or inter-quartile range. (IQR)
- 8. Descriptive Statistics • To illustrate the points, in this topic we will be using a data (Data of a study about hypertension among 150 adults.)
- 9. level of its measurement How to describe a variable
- 10. Analytical tools for more effective research. Data Collected Data filled in SPSS Ready for analysis ??? Check Normality distribution Test for Normal distribution Data Normally distributed Apply Parametric test Data not normally distributed apply Non- Parametric test
- 11. Describe Numerical Data • For such study, normally the first thing to do is to describe the demographic characteristics of the study samples. • Let us start by analyzing the age distribution.
- 12. SPSS Analysis: Describe numerical data 1. Click Analyze 2. Click Descriptive Statistics 3. Click Explore 4. Move age to Dependent List box 5. Click Plots 6. Check Histogram 7. Click Continue 8. Click OK
- 13. Testing for Normality using SPSS- It will help you to determine whether your data is normal or not. Click Analyze > Descriptive Statistics > Explore... on the top
- 14. Testing for Normality using SPSS- It will help you to determine whether your data is normal or not. Click Analyze > Descriptive Statistics > Explore... on the top
- 15. Testing for Normality using SPSS Leave the above options unchanged or Fill Confidence interval as per your study.
- 16. Testing for Normality using SPSS Click Plot the button. Change the options so that you are presented with the following screen: Click the Continue button. Click OK the button. Output: SPSS outputs many table and graphs with this procedure. One of the reasons for this is that the Explore... command is not used only for the testing of normality but in describing data in many different ways.
- 17. SPSS Output • The results here show that the mean (SD) is 37.6 (7.6) years • median (IQR) is 37.5 (10) years • Ranging from 18 to 56 years old. • Skewness and kurtosis are −0.254 and −0.198,
- 18. Results of Normality using SPSS Shapiro-Wilk Test of Normality • The above table presents the results from two well-known tests of normality. • Shapiro-Wilk Test is more appropriate for small sample sizes (more than 50 samples) but can also handle sample sizes as large as 2000.
- 19. Results of Normality using SPSS If the Sig. value of the Shapiro-Wilk Test is greater the 0.05 then the data is normal. Other method is Normal Q-Q Plot
- 20. Results of Normality using SPSS Histogram for Age we might not be able to appreciate its distribution. What we can do is to fit a distribution curve on it. SPSS Analysis: Fitting normal curve 1. At the output window, double click on the graph to open up Chart Editor 2. Check Show Distribution Curve 3. Close the Properties window (You can skip this step) 4. Close Chart editor
- 21. Results of Normality using SPSS • The distribution is a symmetrical bell- shaped, hence age is normally distributed. • Therefore, the age should be described using mean and standard deviation.
- 22. Results of Normality using SPSS Histogram of Age with Normal curve Positively skewed distribution of Glucose
- 24. Describe Categorical Data SPSS Analysis: Frequency 1. Click Analyse 2. Click Descriptive Statistics 3. Move gender to Variable(s) 4. Click OK Result: For categorical data, we use Frequency. Based on Gender table, female 55.3 % (n = 83) of the study population.
- 26. Other tests for analysis Normality distribution Normal Proceed with analysis •Correlation •Regression •Paired Samples t-test •Independent Samples t-test •ANOVA Normality distribution Not Normal Proceed with analysis (time bound studies) Chi Square