Certainly inferential statistics can be applied. That doesn't ensure that the results are valid. In the case of purposive sampling you will have selection bias. In order to improve causal inferences drawn from these data you will need to 'standardize' the responders to the population from whence they were drawn. This is typically accomplished via weighting, where you have characteristics of the population and you first determine whether this group of responders are comparable to the population on those characteristics. From there, you then perform weighting (typically something like "the inverse of the probability of being in the survey") and then you use these weights in all your models.

Without demonstrating that you have balance on important characteristics that may influence how one responds to questions on the survey, you will remain with biased results.

I hope this helps

Ariel

It is possible, but not recommendable for abuse. The reason is that inference is made for a probabilistic sampling design, and purposive design is not a probabilistic sampling design. A consequence of this is that for purposive design it is not possible to obtain the bias, since this requires "to know the expectation of the estimator", but without probabilistic sampling design this is impossible.

The weighting technique, as Ariel suggests, may be used cautiously in some cases. You have to assume that the bias in your sample consists mainly in the proportions in which the units of the universe are represented in the sample. However, if your sample has been taken from just one organization, you would have to assume that this organization has no peculiar characteristics that could influence in a distinctive way the variables you have measured. While your units of analysis are managers, your primary sampling units (PSU) should be organizations. In other words, given the universe of organizations, you have taken a sample of just 1 organization !! Of course, you can use inferential statistics, but in my view, the results can not be generalized. However, you can present the work as an exploratory study. 

Inferential statiscs can be applied with purposive sampling. The issue which need to be considered is the appropriateness of the sample regarding the assumptions of the test which will be used, i.e. normality, linearty of the variables, adequate sample size..etc.

Can you please explain further about normality and linearity of the variables Prof Muayyad?

i recommend inferential statistics

In addition to waht I mentioned above, if you present your questions and the descriptive statistics for the the variables in the study, this will give  a complete clear picture on the suitabilty of inferential statistics, and even what exact test will be used.

Inferential statistics, unlike descriptive statistics, is a study to apply the conclusions that have been obtained from one experimental study to more general populations. This means inferential statistics tries to answer questions about populations and samples that have never been tested in the given experiment.

Inferential statistics infer from the sample to the population, they determine the probability of characteristics of a population based on the characteristics of your sample, they also help assess the strength of the relationship between your independent variables, and your dependent (effect) variables. With inferential statistics, you can take the data from any samples and make generalizations about a population. 

There are two main areas in case of inferential statistics, estimating parameters. This means taking a statistic from your sample data and using it to say something about a population parameter and hypothesis tests, this is where you can use sample data to answer research questions. 

Regards

Gedion Firdawek how did you go about this, I know it is 5 yrs ago it may help me, please email me at [email protected], I am stuck with my proposive sampling data