Month: October 2014

Descriptive and Inferential Statistics: What is the difference?

In a situation when data has to be analysed, like for instance, the marks achieved by a certain number of students for a certain piece of course work, it is possible to put to use descriptive as well as inferential statistics in the analysis of the marks that is to be done. Mostly, in all … Continue reading “Descriptive and Inferential Statistics: What is the difference?”

In a situation when data has to be analysed, like for instance, the marks achieved by a certain number of students for a certain piece of course work, it is possible to put to use descriptive as well as inferential statistics in the analysis of the marks that is to be done. Mostly, in all the researches that are conducted on the groups of people, both descriptive as well as inferential statistics are put to use so as the results can be analysed and conclusions can be drawn. Now, we need to understand what is the real difference between descriptive and inferential statistics?

Descriptive Statistics

It is the term that is given to that type of data analysis which helps to describe, show or help to describe and summarize the data in such a way that meaning is added to it. However, the descriptive statistics do not offer the scope for deriving conclusions beyond the data analysis and help in concluding the acceptance or rejection of the hypotheses. They are only the simpler ways to describe the data. Descriptive statistics play a very important role because if the data is represented otherwise it would be difficult to understand the representation of the data, More, so if there was a lot of data. Generally, there are two types of statistics that are used for describing the data.

1. Measures of Central Tendency: it helps to measure the central position of a frequency distribution for a group of data.
2. Measures of Spread: it helps to measure how spreads out the scores of the data ar

Inferential Statistics

While Descriptive statistics give information about an immediate group of data inferential statistics are those techniques that help us to make use of these samples so as to make generalisations regarding the population from which the samples have been extracted. It is necessary that the sample represents the population accurately. The method for achieving the same is called sampling. There is a fact about sampling that causes the occurrence of Inferential Statistics and that is the occurrence of sampling error and the sample does not represent the population perfectly. The methods of inferential statistics are

1. Estimation of parameters
2. Testing of statistical Hypotheses

A Simplified Comparison of the Key Study Designs

Cross Sectional Study Design: The use of cross sectional design is used for the kind of research that collects data on more relevant variables at a single attempt from different kinds of people, subjects or phenomenon. The key feature here is that the data is collected at the same time.  The cross sectional design offers … Continue reading “A Simplified Comparison of the Key Study Designs”

Cross Sectional Study Design: The use of cross sectional design is used for the kind of research that collects data on more relevant variables at a single attempt from different kinds of people, subjects or phenomenon. The key feature here is that the data is collected at the same time.

 The cross sectional design offers a snapshot of the different variables in the study. It would help to bring forth those variables that are representatives of a cross section of the population. These designs usually depend upon the survey techniques for the collection of data.

Cross Sectional Study Design have key advantages and disadvantages.

The main advantages are:

  •  It collects data on multiple variables
  • Data is collected from a large number of respondents
  • Data is collected from varied subjects
  •  It collects data on attitudes and behaviour
  • It aids exploratory  research
  • It helps to generate relevant hypotheses for future research
  •  The data from this is useful for many different researchers

The main Disadvantages are:

  • Augments chances of error
  • Adds the cost as the subjects increase
  • Increases cost as there are multiple locations
  • Does not measure or evaluate change
  • Does not focus on cause and effect
  • It does not control independent variable
  • It is more time bound

Longitudinal Designs

As the name is indicative, a longitudinal design collects the data over a span of time. The measurement is taken on each and every variable in not just one time frame but over two distinct times. It aids in measuring change and comparing it over time. Time series and panel are two different types of longitudinal designs.

When we talk of the first type, a Time Series Design compiles the data on a similar variable at predefined regular intervals.  This is done in a form of aggregate measures of population. There are a lot of uses of Time Series Designs.

  • Putting up a baseline measure
  • Comparing the changes that have taken place over time
  • Keeping a record of the trends
  • Predicting trends for the time to come

The presentation of Time Series data is in the form of pictorials like charts and graphs. Analysis and interpretation from these graphs looks for four types of patterns that are identified.

  • Long term trends
  • Cyclical Variations
  • Seasonal Variations
  • Irregular Variations

There are some advantages and disadvantages of Longitudinal Designs. These are:

Advantages:

  • It is very easy to collect data
  • Representation of data in graphs is easy
  • Interpretation of data is easy
  • It is useful for prediction of short term trends

Disadvantages:

  • The approach for data collection changes over a time period
  • It is difficult to bring out more than one variable at a certain time
  • It needs to be clubbed with qualitative research for explanations
  • It relies on the assumption that present trends will remain unchanged

Is there a crisis we are facing with smart phone research?

The word “Mobile” is the most commonly used word we see around. It is able to put research in a context which has not been understood before. It provides a whole lot of advantages that researchers have been seeking for ages in order to bring in the accuracy in their analysis and better responsiveness. It … Continue reading “Is there a crisis we are facing with smart phone research?”

The word “Mobile” is the most commonly used word we see around. It is able to put research in a context which has not been understood before. It provides a whole lot of advantages that researchers have been seeking for ages in order to bring in the accuracy in their analysis and better responsiveness. It brings in the idea of just in time feedback and brings in the synergy between the respondent and the researcher even if they are at different locations. Of course, mobile research has grown but it is thought provoking that why it has not boomed?

Research a has revealed that mobile device surveys are on a constant rise but at the same time the dropout rates for smartphones are almost twice. Respondents have been seen using mobile devices as medium for easy participation in research. However, a significant barrier in mobile research completion is the difference in the screen size and the variation in the ease of data entry. It is important in this phase that the screen size, the functions of data entry and the length of the survey should have the flexibility to fit into the situation of the respondent and not just the need of the respondent.

The research fraternity needs to appreciate the ease and comfort with which the respondents use the smart phone devices. In this learning process of smart phone adaptability in research the researchers are learning to overcome the barriers and finding means to squeeze out meaning and value out if shorter mobile texts. There are certain tips that researchers could follow in order to stay ahead in this trend:

  1. Do not resist but respond to change
  2. Customize and offer alternatives for technology involved research by giving options for desktops/mobiles and smart phone alternatives.

The feasibility of technology is all about the tipping points. In time to come, the mobile access of research would actually reach a point where the researchers would have  no choice  but to accept the change. The question that is triggered in each one’s mind is whether research would be able to accept and adapt to the methods and acceptations that are changing or they would remain stuck with the expected but unacceptable dropout rates?

 

Individualities of Destructive and Useful Questions

Individualities of Destructive Questions: Researchers are often found suggesting guidelines for creative research supportive questions. However, before doing that it is important to clearly know, what are the kinds of questions that need to be avoided.  There are certain kinds of questions that can be called as destructive questions for a questionnaire. They have their … Continue reading “Individualities of Destructive and Useful Questions”

Individualities of Destructive Questions:

Researchers are often found suggesting guidelines for creative research supportive questions. However, before doing that it is important to clearly know, what are the kinds of questions that need to be avoided.  There are certain kinds of questions that can be called as destructive questions for a questionnaire. They have their own traits and it is important to identify those traits.

  1. Stay away from questions from basic Yes, No questions. The reason being that they offer a very little understanding of the direction these questions take the research into. The focus of the researcher should be to pink up questions that begin with interrogative words such as, What, How, When, Where.
  2. Questions should avoid the use of any leading terminology. They take the response of the respondent in a predefined direction and are often taken in the category of being manipulative or dishonest questions.
  3. Do not have too many questions that begin with, “Why”. These questions bring up a feeling of defensiveness in the respondent and they may get offended that their actions need to be justified to the researcher.

 Individualities of Useful Questions:

These set of guidelines may be helpful in creating a more responsive, analytical and fair questionnaire.

  1. Incorporate open ended questions wherever it is possible. These kinds of questions go beyond the conventional yes or no. The advantage is that they generate a thinking process for the respondent and keep his focus on the questionnaire.
  2. In the case of an interview session, with open ended questions, do ask clarifying questions so as to get an understanding of the bottom line.
  3. Try to bring in questions that give an understanding about the perspectives, assumptions and actions of the respondent.
  4. Ask for help and ideas. It can serve as a powerful tool when enough faith is shown in peer or at times even the respondent when you ask for help. It may help you to get a fresh insight into the research. You could get clues by putting up enquiries such as, “What questions should I be asking now?” or “What else can I know from you?”