I found regression models in many fields to measure variables but I am wondering about the methodology to develop a prediction models.
Regression models are predictive tools. If you want to use them to predict the behavior of an individual, you must set "behavior" as your dependent variable, and you must have indicators and measures for such. You must also identify your explanatory or independent variable likewise you must have indicators and measures for such.
If you have built your model on the basis of collected data, you will be able to predict the value of behavior given a value of the explanatory variable.
What type of behavior are you looking at? If you want to predict how someone will vote in an election, you will use a nominal logistic regression model. If you want to see if someone will spend a certain amount of money in a given day/week/month/year, you need to run a poisson regression. If you want to predict the performance of a stock in the stock market, you will use forecasting and time series data.
thanks a lot professor Ekstrom. I am looking for predicting cheating behavior
That would be a standard Logistic Regression. How many factors do you want to test? How much data do you have already?
Hi, I am running a study on predicting behavioural intention towards cloud computing adoption within the financial sector. I have already used the PCA to compute my constructs;
1) perceived benefits, perceived ease of use, perceived benefits
2) perceived costs and perceived risks.
I then used compute variable function to compute the means of the these factors under mean_pcr and mean_peub
I have collected information regarding firm size, industry type, work experience within the firm. However, I was wondering how can I predict the behaviour of the sample towards these two factors that I just mentioned. I want to test my hypothesis which is;
- perceived ease of use, perceived benefits and perceived usefulness will negatively impact the behavioural intention towards cloud computing usage.
previous studies like tripathi (2017) have used hierarchal multiple regression to test the following hypothesis, however I am not able to understand how can I use this regression; like which factors am I supposed to input into the regression function to predict behaviour.
moreover, which regression would be possible for this type of a case?