Hi Everyone,
I’ve not posted for more than 2 weeks…I understand if I need to keep my blog active I need to post regularly atleast once a week :D.
But I strongly feel I should post only If I have something valuable to share with you all……
I had already written about research in my previous posts and published mynotes.
But it is not an easy unit to understand for most of us.
In the past weeks,few of you had some doubts regarding statistics & research.I thought why not explain the stuff in a simple language.
So here we go…
DATA TYPES:
a)NOMINAL SCALE: The lowest level of measurement in statistics.The data is classified into groups on the basis of gender,names etc.
You can count the number of males/females or names but cannot measure the gender or names.Other examples:blood type,hand preference etc…
Nominal data can be compared with Chi-Square test and the central tendency can be measured with Mode..
b)ORDINAL SCALE:This data can be ranked and ordered.It still can’t be measured.The subjects/objects are classified based on the degree to which they possess the characteristic. eg:Manual Muscle Testing,Sensation,Pain scales.Surveys often create ordinal scales to find out people’s preferences say on a scale of 1-5 or most preferred-least preferred.
There is no true “zero point” for this data.It is chosen arbitrarily.Ordinal data can be correlated with Spearman’s Rank order or Kendall’s tau rank correlation.Their central tendency can be measured with Median or Mode.
c)INTERVAL SCALE:The data is classified based on predetermined equal intervals.Scores can be added or subtracted but not multiplied/divided.Eg:(We could say the difference between 20-30* and 90-100* is 10* but there may be a difference in heat)Temperature scales,Calendars,IQ charts.
There is no “true zero point” for this data too.The correlation for this data can be measured with Pearson’s product moment coefficient,Multiple Regression etc., and central tendency with Mean,Median or Mode.
d)RATIO SCALE:The most precise/highest level of measurement wherein you can count and measure data with “true zero point”.The data has equal intervals.Ratio scales are used to gather quantitative info like surveys asking respondents for age,income etc.So an 80 year old person is going to be twice as old as 40 year old.eg:Height,Weight,Range of Motion Scales,Annual Income etc..
The central tendency can be expressed for ratio data with Mean,Median or Mode
CENTRAL TENDENCY:This is the way in which you would summarize the above 4 types of data,yet still retain the necessary info.
There are three types namely,
1)MODE:The most frequently occurring value in a set of data.
For example let’s take a set of numbers:94,45,56,68,79,30,40,38,45
In this 45 is the mode,the most frequently occurring number.
2)MEDIAN:The middle score in a set of data
Take the above example and re-arrange in ascending order: 30,38,40,45,45,56,68,79,94
The median would be 45 leaving 4 values above and below.
In case of even number of data say:30,38,40,45,45,56then median would be midway between 40 & 45 i.e 40+45 divided by 2.
So it would be 85/2= 42.5
3)MEAN:The arithmetic average of all the data.
The mean(M) is the sum of a set of scores(X) divided by the number of scores(n).
M=X(x1+x2+x3+……xn)/n
For example take a set of numbers:10,20,30,40,50
Here X=10+20+30+40+50=150 and n=5
Then M=150/5= 30
SKEWNESS:Asymmetry in the distribution of values is called skewness.In a normal distribution the mean,median and mode all coincide.But in a skewed distribution it is not so.
In a positively skewed distribution(skewed to the right),the mean>mode and mode>mean in negatively skewed(skewed to the left) distribution
That’s it for now.Hope to add more info on Statistics in my next post.
physicaltherapyprep 