Dear Osmar,

I'm PhD in Psychology at University of São Paulo (USP) and now I'm a Post-Doctorate Researcher at University os Campinas (UNICAMP). In both I worked with qualitative research.

I think that in the situation you describe it depends on the scientific method you adopted.

If you adopted a method of participatory monitoring (as I did in my post-doctorate research), there is an interference of the researcher. In my research I worked with children with hydrocephalus who could not express themselves, and there is a need of interference of the researcher to help them to find a way, to build a means to express themselves.

On the other hand, if you adopt a method where you just observe the result (as in my doctoral research), your participation as a researcher will be retrospectively and theoretical. In this case I worked with Discourse Analysis (Pêcheux) and data were analyzed taking into account theory and socio-historical context.

Best regards,

Dionéia

I certainly think that if respondents are only providing answers to the questionnaire to avoid future reminders, then you will certainly get different results, in a variety of ways, than you would if you have interested respondents who are engaged in the topic of the questionnaire. That, of course, is also interesting data. Perhaps that kind of response data from potential participants might be an impetus to revise the questionnaire or to seek out additional sampling techniques that might be successful. It also depends on what your sample selection criteria are.

In qualitative inquiry, we are typically not that concerned with obtaining a large enough sample to achieve statistical significance. That's more of a goal for positivist research with surveys. So you might also consider the goals of your research and how important a large sample is based on the theories that guide your work and how your methodology was produced from that. It might help you view what now seems like a concern and flaw in research design in a different way. I hope that helps!

Dear Jessica,

thank you for your contribution!

Best regards.

When there is a low level of interest on the part of the participants, you will have a hard time getting responses to mail out questionnaires for both qualitative and quantitative data. In either case, the amount of effort required is likely to be a factor, so the more work the person has to put into responding, the less likely that you will get a reply.

As Jessica notes, qualitative researchers typically work with smaller Ns, but part of the reason for that is the use of purposeful sampling where the goal is to select people who are closely involved with the topic, and then to hear from them in depth and detail.

The insistence of the researcher to get answer may be 'embarrassing" for the respondent who may say anything just to 'please' (social desirability' or to 'get rid of' the researcher. So, the reliability of the responses is jeopardized.   

In addition to what has been said, I would like to add that qualitative research tend to generalise the outcome to the phenomenon rather than to the population - which is the case of quantitative research. Regarding to your question, I would discourage you to be too insistent when trying to gather responses due to ethical reasons: it is very important that participation in your research is voluntary; and, therefore, you should take precautions to not make your participants feel coerced to take part in your research.

Dear  David, Hamida and Claudia,

Thank you for your contributions! Indeed, the insistence on searching for answers involves ethical issues and can be embarrassing to the respondents. 

This has been shown frequently, when a large number of studies are conducted with the same population, as is the case currently with the managers of PMEs in Brazil.

Greetings!

Dear Osmar,

You're very welcome! In my research group we try to overcome these difficulties by offering some type of financial incentive for answering our surveys, but I agree that it's not easy to get respondents. Within the social sciences, we're constantly dealing with this issue.

Good luck with your research!

Dear Claudia,

Very grateful for your message! I am directing three studies in higher education and on the relationship between the accounting professional and their clients, as well as the preparation of to the current challenges of the profession. I understand the anguish for the students, opposite the small number of questionnaires that are answered and in the current environment is difficult to give some encouragement to get more answers, because there are no funds for that.

 My best regards!

These are all good answers. I don't know whether this has been other experience, but I have found it is difficult to recruit participants if they are not interested in the research topic or are interested in the research topic, but worry about the findings.  For example I recently read a  dissertation on spinal cord injuries seeking patient, family, and professional views.   The institutions from which the  PhD student was going to seek professional opinions refused to let their professionals participate.  Studying client /professional accounting challenges could be part of students anguish. 

Dear Osmar,

Your raised undoubtly a very interesting topic. I have personally made a few research with questionnaires online and have the opinion that you have to send some reminders periodically in order to get a reasonable amount of answers. My understanting is that this is depending on a number of issues: the type of questionnarie, the target of the research, the profile of the respondents, etc. But I think there is another effect related with the "timing", that is, which day you send it, at what time and how long it takes to complete the survey (specially if you are focusing on working respondents) . It might be that the respondent wants to answer but maybe they flag the message and they just forget about it. Therefore I think there nothing wrong with sending a few reminders (up to 3 or 4). It also works to state a final date for gathering the answers.

Sometimes it is not only the small compensation token that works, but the commitment to send the results report, but this clearly depend on the topic of the research and how interesting for the responden it might be.

Best regards,

Javier C.

Dear Osmar,

I'm PhD in Psychology at University of São Paulo (USP) and now I'm a Post-Doctorate Researcher at University os Campinas (UNICAMP). In both I worked with qualitative research.

I think that in the situation you describe it depends on the scientific method you adopted.

If you adopted a method of participatory monitoring (as I did in my post-doctorate research), there is an interference of the researcher. In my research I worked with children with hydrocephalus who could not express themselves, and there is a need of interference of the researcher to help them to find a way, to build a means to express themselves.

On the other hand, if you adopt a method where you just observe the result (as in my doctoral research), your participation as a researcher will be retrospectively and theoretical. In this case I worked with Discourse Analysis (Pêcheux) and data were analyzed taking into account theory and socio-historical context.

Best regards,

Dionéia

I think the number of times resending the same questionnaire to the same respondents should be limited so that you do not receive responses from less-interested respondents. More importantly, I would focus on the way that I put my questions as well as on the wording of such questions in order to limit the hasty answers from those respondents.

In qualitative studies, the researcher's interest lies with the evaluation of the phenomenon through various types of evidence. The questionnaire is one element among others to be considered; so there isn´t minimal or valid number, as worth the criterion of saturation of information. In my research in Accounting, I use the Book of Bob Ryan, Robert Scapens and Michael Theobald [Research Method & Methodology in Finance and Accounting], especially chapter 8.

Dear Claudia and Dionéia, 

Thank you for your contributions!

 Dear Ahmed,

Thank you for your contribution, notably the suggestion to limit the number of tantativas in getting answers. Greetings!

Dr. Lavarda,

Very grateful to contribute. Indeed, the suggested book will be of much use. 

Best regards!

Prezado Prof. Osmar

Realmente a insistência para receber o retorno de um questionário de pesquisa pode de certa forma comprometer o estudo, mas tudo isso depende do posicionamento ético do respondente. A sobrecarga muitas vezes de atividades profissionais pode influenciar o respondente depois de muita insistência a responde de maneira superficial a pesquisa. Sabemos da necessidade e importância que estes estudos aplicados têm para a pesquisa, acredito ser a melhor forma de validar e consolidar o conhecimento.

Dear Osmar,

We work with discourse analysis and when we find data that we didn't expect, these are taken as something that can give a new direction to research. Michel Pêcheux ("Discourse: structure or evenement?") calls this phenomenon as an evenement. Under the application of the theory of Discourse Analysis, we observe that the cohesive image of subject described by the scientific discourse is broken with the emergence of a split subject, constituted on a discourse of another order, outside the scientific discourse logic, and that "displacement" of the rigid structure of the science characterizes the Decisive Moment (an evenement) as a resistance to the dominant ideological formation of this discourse.

Thank you for participating in this important theoretical discussion!

Best regards,

Dionéia

Thanks to Alzenir and Dionéia for the valuable contributions!

Hello everyone! I greatly appreciate the contributions. In particular, the reflections sent by Sharon Anderson and Javier Chamacho allow the understanding that the difficulties of investigating are not unique to one or another area. 

Greetings!

Thanks Osmar,  I think the difficulties of recruiting research participants are not unique to qualitative research. Quantitative research has similar problems. In health research they will pay recruiters to sit at the emerg desk or GP's to recruit for them. I reviewed a paper yesterday where they decided to recruit anonymously for survey completed on Survey Monkey. I certainly had some misgivings about the sample.  Not clear what % of the population participated nor (because it was self-report) that they really met the criteria.  

Dear Osmar,

I just remembered Henry&Moscovici, that I quoted in my PhD thesis talking about the difficulty of theoretical model:

(French and Portuguese versions)

Segundo Henry e Moscovici (1968), a análise de conteúdo “não conduz um estudo da linguagem e nem de sua manifestação enquanto tal”, mas leva à “determinação mais ou menos parcial das chamadas ‘condições de produção’, pois se utilizam de um aparato conceitual pré-existente que aprisiona o analista dentro de um modelo teórico, impedindo-o de estudar outras variáveis, fechando sua interpretação” (LAUREANO, 2005, 10):

Nous sommes maintenant em mesure d’enoncer le problème fundamental de l’analyse de contenu. Il est clair que les deux plans de réference que nous venons de definir ne sont pas independants. C’est ce qui simultanément rend analyse possible et fait le problem. [...] En effet, les conditions de production sont caracterisées à partir de elements dintigués dans le texte et les possibilités de cette caracterisation dependent de leur definition. En l’absence de definition autonome de la structure de ces deux plans et de leurs rapports, on about à une impasse methodologique et theorique. Tous les problèmes de l’analyse de contenu dérivent de celui-ci. Les pratiques analytiques ne sont que des moyens d’échapper à la circularité de cette problématique sans pour autant la resoudre  (HENRY & MOSCOVICI, 1968, p. 38, apud LAUREANO, 2005, p. 10-13)[1]

[1] “Nós estamos agora no ponto de enunciar o problema fundamental da análise de conteúdo. É evidente que os dois planos de referência que definimos não são independentes. É isto que simultaneamente torna a análise possível e lhe cria problema [...] Com efeito, as condições de produção são caracterizadas a partir de elementos distintos no texto e as possibilidades desta caracterização dependem de sua definição. Na falta de uma definição autônoma da estrutura desses dois planos e de suas relações, atinge-se um impasse metodológico e teórico.

Todos os problemas da análise de conteúdo derivam disto. As práticas são somente meios de escapar da circularidade dessa problemática sem, no entanto, resolvê-la” (trad. de LAUREANO, 2005, p. 12)

Referências

HENRY, P.; MOSCOVICI, S., Problèmes de l’analyse de contenu, In Langages, n. II, sept. 1968.

LAUREANO, M., A interpretação (revelar e esconder sentidos): articulações entre análise do discurso e psicanálise lacaniana, Tese de Doutorado, Programa de Pós-Graduação em Psicologia, FFCLRP-USP, Ribeirão Preto, 2005.

Hello everyone,

I reiterate my thanks to everyone who contributed to the questioning. The set of perceptions, along with the experiences I had in two investigations, resulted a technical and opinionated text, published in Accounting Journal, which is available in my contributions here at RG.  Also, I am designing a study with greater depth on the subject .

Best wishes to all !