Hi Researchers,
Dear scholar I'm facing problem regarding selecting techniques for data analysis of my dissertation. The classification of my model is one IV, one mediator and one DV. Independent variable was industry (importing and exporting businesses) whose data was collected as binary number 0,1. The data of mediator (strategic typology) and dependent variable (innovative business model) were collected using 1-5 Likert scale. Now the problem is how can I test the direct and indirect (mediation ) relationship between variables. Specially when IV was in binary number. Kindly share your valuable suggestions? IF anyone share link of technique or research paper using that particular technique that will be very helpful. Thank you.
Evalutation plan for business modell adherent strategies of import and export: Beam |/O to success by power of strategy.
1. DV= f(IV) = Binary functin of IV - preleminary analysis of a fantastic relationship.
2. Differences of DV depending on | or 0: student's test.
3. Modificaton of DV by mediator: chi-square test for distribution, independency and homogeneity of varicance.
4. Inversed modificaton of Mediator by DV within matched data sets: Check wether the original matrix of rows and columns shwos linear dependency for square matrixes of IV and any pair of DV. By this You can find out those strategies, that have strong influence on the business model. Call them business model adherent strategies.
Strategic type and business modell should show a strong connection with exclusion of inappropriate matching.
Propability of connection and artefacts
You find a table with enlarged overview about relationships in one of my German concept papers in the journal IZPP.
Present Your supposed model in Your thesis. Give reasons for a refusal of an not appropriate model as the proprosed binary function of the independent variable.
This will give Your paper as well a proper design as well a laconic analyis. Hope to read Your thesis soon.
I am not sure if I understand what you are doing, but it appears that you are predicting an ordinal outcome variable, and ordinal mediator or covariate, and a binary group variable. It appears that you are not interested in the interaction, and you haven't mentioned any worries about non-mediator paths in the causal models. If you were assuming the 1-5 scales were interval, would you just be doing an Ancova (in steps if you wanted all those statistics folks who use the word "mediation" often report), so is your question what to do because you want to treat these variables as ordinal or is your question asking something else. Be best also to look at http://davidakenny.net/cm/mediate.htm and the sources cited there.
@Daniel wright thank you. Your given link was very helpful and i find the required information. One of the link on this page shows that you can assign labels in values in SPSS to your dicotomous IV and rest is fine.
@Martin Wedig can you please provide link of your paper or its name so that i can read it.
@ Syed Shahbaz ul Hasna, You find a list of explanations in table 4 in
Article Leben – eine Annäherung vom Tod her.
- Badges
- Science topic