Home › Forums › Methodspace discussion › Descriptive Statistics
- This topic has 4 replies, 4 voices, and was last updated 10 years, 9 months ago by
Deni sapta Nugraha.
-
AuthorPosts
-
29th March 2010 at 4:08 pm #4846
Rudy Renteria
MemberWhat exactly is descriptive statistics and regresssion analysis?
1st April 2010 at 3:27 pm #4850Deni sapta Nugraha
ParticipantDear Rudy,
I would like to quote the definition and the illustration about your question from the encyclopedia of applied linguistics, hopefully it helps you to define what you are questioned.Descriptive statistics is statistical procedures that are used to describe, organize and summarize the important general characteristics of a set of data. A descriptive statistic is a number that represents some feature of the data, such as measures of CENTRAL TENDENCY ( any estimate of the central point around which scores tend to cluster. The most common measures of central tendency are the MEAN, the MEDIAN, and the MODE) and DISPERSION (the amount of spread among the scores in a group. For example, if the scores of students on a test were widely spread from low, middle to high, the scores would be said to have a large dispersion. Some common statistical measures of dispersion are VARIANCE, STANDARD DEVIATION, and RANGE) see Richard and Schmidt (2002).
Regression Analysis is a statistical technique for estimating or predicting a value for a DEPENDENT VARIABLE from a set of INDEPENDENT Variables. For example, if a student scored 60% on a test of reading comprehension and 70% in a grammar test (the independent variables), regression analysis could be used to predict his or her likely score on a test of language proficiency (the dependent variable). When two or more independent variables are present, as in this example, the statistical technique is called multiple regression. see Richard and Schmidt (2002).
1st April 2010 at 4:26 pm #4849Rudy Renteria
MemberThank you, this was is very helpful….
5th April 2010 at 11:46 am #4848Steve Moran
MemberHi Ruby
descriptive statistics…these are the likes of means & Standard deviations coupled with histograms etc.
the purpose is to illustrate the behaviour of the phenomenon you are interested in and how it is likely to behave from which you can extract probabilities.
Regression
This is when you are interested in relationships between two variables where it might be one depends on another.
Usually the data(X,Y) are plotted on a graph known as a scatter graph. The question is can you plot a line through the middle so that this “line” can be used as a realistic representation of “how well Y depends on X” subject to variances.
Techniques exist to get the “line of best fit” or line of regression.Usually to add quality to your appraisal of the data there is a statistic called a correlation coefficient. This is used to describe how tight or close the scattered data is to the line you have estimated as reasonable representation of the relationship between Y and X. It also helps decide if there is or is not a relationship at all.
16th April 2010 at 5:59 pm #4847varadi vijay
MemberDescriptives : Mean, Median, Mode – will give properties of the data
Standard Deviation, Variance – Will explain the variance between the data
Skewness, Kurtosis – How the data behavesRegression : A technique used to discover a mathematical relationship between two variables using a set of individual data points.
-
AuthorPosts
- You must be logged in to reply to this topic.