21st July 2010 at 5:12 pm #4326AFHParticipant
Can I use inductive approach in quantitative research? Can I solve my research questions directly with statistical anlysis but without testing any hypotheses? Please kindly advise.22nd July 2010 at 10:15 am #4333Sharon KibbleMember
HI I am currently using a quantitative approach to discover which inductive codes have increased or decreased in a series of findings from various medical students. So as far as I am aware there are a number of questions you need to answer to decide what are the right data analysis tools for your research for example:
1) What are your research questions? 2) Do any of them have a directionality to them ? for example are you expecting something to increase or decrease after your investigation?
I have found so far in my research that many people think there is no hypothesis in qualitative research but I always think that if I have a question then in a way it can also be viewed also a hypothesis. Please let me know if you would like to discuss more.22nd July 2010 at 10:44 am #4332AFHParticipant
Great! Thanks so much Sharon.
I am using multiple regression to analyse my set of data that was acquired from a survey, so that the relationships between my indepednent and dependent variables can be shown. I would like to use the quantitative findings obtained from my multiple regression analysis on answering my research questions.
So far, I am not able to find any literature to justify this move. Usually quantitative research adopts deductive approach and connects to hypotheses.
Please kindly advise.22nd July 2010 at 11:39 am #4331Sharon KibbleMember
I too used Regression analysis in my BSc thesis to look at the predictor variables of age and gender and length of service on the scores of a survey I had conducted. This was approved by my University tutor at the time and I was awarded a 1st for my dissertation. So I think you are going along the right lines.
A good book to view regression analysis and its usage in the 1st instance is Andy Fields “Discovering Statistics in SPSS. Are you conducting your research for a university? If so then maybe the statistics tutor there or a researcher can advise. Justification can not always be found in the Literature but as long as you know that you have predictor variables that may be influencing a survey results then regression may be a good way of analysing results.25th July 2010 at 11:22 am #4330Abeya El BakryParticipant
I don’t know much about statistical analysis, but if you don’t test the hypothesis, how are you going to verify your data. Ultimately what you need is to validate your data and relate it to your research questions and literature review, so how can you do that without testing your hypothesis, or relevant assumptions.
The inductive approach is useful in quantifying the issues that you have explored, it will not give you the reasoning behind it, but it will show you the apparent factors and their relationship to each other, that should be useful.3rd September 2013 at 9:02 am #4329Saleh AliParticipant
You are right!
in case study approach,for instance, we have already some assumptions. As one of my supervisors keeps saying : we do not enter the field with “empty heads” but rather with “open minds” (Cathy Urquhart Grounded Theory For Qualitative Research: A Practical Guide)
The challenge, however, is not to pose our preconceptions.17th September 2013 at 2:30 am #4328Ben KohParticipant
In general, inductive reasoning is associated with qualitative approaches, while deductive for quantitative. But research is never cut-and-dry. It depends on the issue in question, or in simple terms: what is it you want to know.
If you begin with an inductive approach, what you are doing is exploring the field for answers “from the ground up”. You let the field guide you and have an open mind. Although open-ended or semi-structured interviews are the most common tool, a survey can also be used. But the survey design needs to allow open answers (ie the response cannot be restricted to yes/no); words such as “please explain”, “elaborate what you mean” would be common terms used.
Whilst not my personal preference, once these qualitative data is obtained, you can start “coding” them into quantitative operational terms. For example, if your qualitative survey asked “explain to me some of the influences that would make you buy a car”, and some of the answers you get (from a long and elaborate interview transcript) included “if it was red in colour”, “if it matches my handbag”, “if the colour does not get dirty” etc, then you can code those interviews with “colour as a reason”.
So what you have done is convert a qualitative data into a quantitative one “colour”. You can then compare it with another converted code e.g. Bought car yes/no.
By having a frequency (no. of cases that indicated colour as a reason; ordinal data) v bought car (nominal data yes/no) you can then perform a ANOVA.
If you show a significant difference in the ANOVA, then you can do a follow-up deductive study (from top down). Your survey will then have a closed-ended list (e.g. Is colour a reason why you would buy a car: yes/no). Your hypothesis would then be: colour is a reason whether a person will/will not buy a car.
Hope that helps.26th September 2013 at 10:50 pm #4327Saleh AliParticipant
I found a good article titled ”INTEGRATING DEDUCTIVE AND INDUCTIVE APPROACHES
IN A STUDY OF NEW VENTURES AND CUSTOMER PERCEIVED RISK” This article discusses combining inductive with deductive reasoning together. It suggests that most of the researchers use mixed of these two reasoning but they might not state this realize it! This might not answer your question directly but you might find it useful.
Wish you the best
- The forum ‘Default Forum’ is closed to new topics and replies.