Institutional vs individual unit of analysis

Home Forums Default Forum Institutional vs individual unit of analysis

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
  • #4101
    Alexia Haskins

    Hi there

    I am conducting research that requires me to collect information on perceptions both at an individual and institutional level but i am having a hard time validating the institutional responses as they will be from individuals in an institution….ie i want to know institutional perceptions on a concept and then individual ones. Has anybody conducted a similar study or know of any i could use as an example.
    Would my unit of observation (for the institutional survey) be the individual but my unit of analysis the institution?

    Many thanks!

    Bex Hewett

    Hi Alexia

    It’s difficult to answer fully without knowing your exact subject of enquiry and it the answer also depends on whether you are collecting qual or quant data but for many variables you could say that the institution is reflected by the individuals in it so asking the same questions of a number of key individuals (senior managers/board members etc) might give you some sort of representative sample. You might also want to consult some secondary sources (company website, annual report etc) for the “official” view but, again going to depend on what you are studying. When I did my Masters thesis a few years back I collected qual data from senior managers and from secondary sources and used textual analysis to build up a picture of the concepts that I was studying.

    Not sure whether that’s been helpful…good luck!



    What do you mean by ‘information on perceptions’? Is it just like interview or survey?

    Alexia Haskins

    Hi Bex

    Thank you for getting back to me- to clarify a bit more, i am interested in looking at the concept of ‘ecosystem services’ and how, since its inception, it has been used to guide environmental management and socio-ecological research to reconnect humans and the environment, and whether institutions that use it perceive it to be a useful tool or not. The data will be a combination of qualitative and quantitative and i plan on using an online questionnaire (to gain the institutional perspective) and also do more focussed semi-stuctured interviews.
    I am definitely also planning on looking secondary data from institutions, but your suggestion of asking a few individuals from each institution is really helpful.

    Thanks again

    Alexia Haskins

    Hi Ramanath

    I plan on using an online questionnaire made up of structured and semi-structured questions, and also do follow up, more in-depth and focussed semi-structured/open ended interviews.
    Hope this clears things a bit!

    Many thanks


    I found your question intriguing because it is a common one for beginning researchers to ask. Perceptions are always formed and obtained from individuals. Perceptions solecited are usually about something external to the informant and often about much larger enviironmental phenomena (organizations, institutions, social classes, relgiions, political parties, etc). If your theory, inferences, or deductions are about the institution, then that is your unit of analysis. Summarizing or mathemetically aggregating individual observations about an institution still means the analytic level is institutional.

    Conor Horan

    Hi Alexia

    If I understand what you’re interested in is actually ‘levels of anlaysis’ as opposed to ‘units of analysis’ – studying a phenonmenon from the persepctiv eof the human actor would represent the specific daya to day experiences of actors as they see it and live it tend sot be more subjective in nature and lends itself to collecting qualitative data as meaning become simportant. If you study the same phenonenon from a group or institutional ‘level’ the perspective you get back when data collecting might be more generalist in nature and they issues raised might be more macro based trends. Thus the data could be more quanititave in nature or might be found in secondary sources that have gathered data using surveys for example – this data might inform policy more.


    If you want to justify this approach you can find it in research that discusses ‘levels of anlaysis’ or ‘multi-levelled anlaysis’ (do note that many IT MIS areas use this term in a different way). Case study methodology also supports a mulit-levelled approach to understanding a case.    

    Jeremy Miles

    How this is done depends on what you are doing.  It sounds like you might want to do some sort of multilevel modeling.  In multilevel models, you have units at two levels.  People are level 1, and you measure some things at this level.


    Then you have organizations at level 2, and you measure other things at this level.  So, for example, you might have large organizations and small organizations, and you want to see if happiness of the people differs between them.  For this, you would use a multilevel model.



    Dr, Vinod Sen

    hi dear

    questionnaire should be short and simple, please give some choice to the respondent so that they don’t have to spent time to write the answers. 

    with best wishes


Viewing 9 posts - 1 through 9 (of 9 total)
  • The forum ‘Default Forum’ is closed to new topics and replies.