Notes on various papers all labeling themselves as a multiple-case study. Margaret Burnett, 4/16/07. Bottom line -- multiple vs. single seems to be a matter of degree. A case study involving just one "case" is single -- if it involves multiple, there is more than one "case". Think of a "case" as a situation. So it is more than 1 case/situation if the 2 things are sufficiently different to say it's more than 1 situation. OBVIOUSLY SINGLE: 1 subject who is a juvenile delinquent (JV) is a single-case study. GRAY AREA: 3 subjects who are JVs might be one (researcher views these 3 guys as having enough in common with the setup of the RQ to constitute a single case w 3 observation units), or maybe as being multiple cases (an mid-teen Asian male, a young male WASP, and an almost-20 female). ("In the CASE of the Asian male...). OBVIOUSLY MULTIPLE: 3 sortof-alike male JVs in Portland, 3 in sortof- alike male JVs in LA, and 3 sortof-alike male JVs in NY. Note about context -- anything "bigger" than the case can still contain influential factors, but it's part of "context". How to determine where the context stops and the case starts -- consider the RQ, the case is often mentioned in the RQ. If not, the case is the smallest possible thing that you think still makes sense as the "unit of analysis". Excerpts below from various case studies that all describe themselves to be multiple: Example: 15 families: 5 representing one case, 10 representing another case: ------------------------------------------------------------------------ ------------------------- Goes on to different ways of splitting the cases too below... > The purpose of this study was to describe the similarities and > differences of a small group of families of normal private > preparatory school adolescents and a small group of families of > adolescents with diagnosed substance-abuse disorders with regard > to: level of marital adjustment, family roles and the degree of > family satisfaction, and problem-solving skills and communication > styles. In this study, the problem-solving abilities of the > families were observed and described with regard to flexibility, > ability to resolve issues, clarity of communication, and verbal > communication styles in an effort to explore possible differences > between these two groups of families. > > In this multiple-case study, three research questions were > addressed by a number of measures which included structured > interviews, assessment questionnaires, and direct observation using > a videotaped problem-solving exercise. The questions addressed in > this study were: (1) Do trends in the data suggest any differences > in the level of the marital adjustment between the parents in the > two groups? (2) Do the data show trends suggesting that perception > of roles and the degree of family satisfaction felt by the > individual family members might be significant indicators of family > functioning? (3) Do the data suggest any differences in the two > family groups' problem-solving skills and styles of communication? > > Subjects > > Fifteen families of adolescents were included in this study; five > were families of substance abusing/mood disordered adolescents, and > ten were families of normal private preparatory school adolescents. > For the purpose of this study, normal private preparatory school > adolescents were defined as adolescents who had not been involved > in therapy for a clinical or substance-abuse disorder and who were > currently succeeding, both academically and socially, as evidenced > by average or above average achievement and participation in one or > more major extracurricular activities. Hereinafter they are > referred to as the "normal" group. Substance abusing/mood > disordered adolescents were defined as adolescents clinically > diagnosed as exhibiting a substance-abuse disorder and a mood > disorder at the time of this study. They are referred to as the > SAMD group. > > Each family in the study consisted of a two-parent family. Of the > ten families of normal adolescents, two represented atypical units > in which the father was a stepparent, while the other eight family > units were comprised of natural parents. The five families of the > SAMD adolescents included three natural parent units, one adoptive > family unit, and one family unit in which the mother was a stepparent. > > All ten of the normal adolescents attended a college-preparatory > private school in Southern California. Each of the five SABD > adolescents were engaged in some form of clinical treatment: three > were inpatients at a private psychiatric hospital in Southern > California; one had been a patient at the same hospital, but was > transitioning from a residential treatment program to outpatient > therapy; and one was an outpatient in treatment with a private > therapist. All five adolescents had also been given a dual > diagnosis of a mood disorder by the primary psychotherapist. > > ... > > DISCUSSION > > This multiple-case study presented the opportunity to examine > fifteen families in three distinct areas: level of marital > adjustment of spouses; degree of family health or dysfunction as > perceived by family members; and communication and problem-solving > skills of each family member as evaluated in dyadic assessments. > Although unable to imply that any of the characteristics noted in > this study represent definite distinctions between families of > normal adolescents and families of SABD adolescents, several trends > were noted with respect to the three areas addressed. > * Example: 5 internet courses: ----------------------------------------- Each course was a case, > > We conducted a multiple-case study of exemplary Internet courses, > guided by two initial research questions: > > 1. What are the elements and attributes of current exemplary > Internet courses? Are there common construction, design, > application, and interaction elements in Internet courses of > excellence? > 2. What is the role of learning theory, if any, in current > exemplary Internet course development? > > Methodology > ... > > The first phase in the formal study involved identifying exemplary > Internet courses using the reputational case selection method of > expert referral (LeCompte & Goetz, 1982) . Instructional > technology subject matter experts (SMEs) were queried by email for > attributes of exemplary online courses. Their responses were > compiled and integrated to create an exemplar profile used to > screen course nominations. > > Having created a screening profile to identify an exemplary > Internet course, nominations for exemplars for study were solicited > from professional listservs (instructional technology, distance > learning, and education), organizations that have evaluated > Internet courses for awards and recognition, professional > colleagues, and authors of current literature on online teaching. > > The setting for this study was higher education in the English > speaking world. A cluster of five courses was selected from a > nomination pool of about 70, and studied concurrently. This method > resulted in a highly efficient and uninterrupted flow of data, with > the responses of one developer filling the lag time of another. > This study aimed to look beyond the predefined attributes > qualifying the courses as exemplars, for deeper and unforeseen > insights on excellent Internet courses. Five sources of data were > included in the study: (1) documents, such as course syllabi and > online course materials, (2) archival records, such as threaded > discussion board records, (3) interviews conducted by email, in > person, and by telephone, (4) direct observation, including the > researcherÕs narrative log, journals of observations and > reflections, and (5) artifacts such as student projects and papers. > Data were analyzed and coded using qualitative research and case > study conventions (Miles & Huberman, 1994; Stake, 1995; Yin, > 1994) . One course was dropped from the study, as major course > elements were found to be incomplete. Data saturation was reached > with the remaining four courses in the study. > Example: 4 companies' use of use-cases. ------------------------------------------------------- Each company was a case. > B. C. D. Anda, H. C. Benestad, and S. E. Hove (2005) > A Multiple-Case Study of Effort Estimation based on Use Case Points > In ISESE'2005 (Fourth International Symposium on Empirical Software > Engineering). IEEE Computer Society, Noosa, Australia, November > 17-18, pages 407Ð416 > > Abstract > There is much interest in industry for the estimation of software > development effort based on use cases, but little scientific > evaluation of applying use cases in estimation has been reported. > We investigated one particular method, the use case points method, > in a multiple-case study. The Software Engineering Department at > Simula issued a tender for a system, and 35 companies responded, > with estimates ranging from 78 to 654 hours. We chose four > companies to develop the system. They all implemented the same > functionality, but their development processes varied, ranging from > a light, mainly code-and-fix process, to a heavy process with much > emphasis on analysis and design. The use case points method > estimated this project to 430 hours. This was equal to the actual > effort spent on implementing the system by the company with the > lightest development process. In our opinion, the results from this > study may represent a basis for measuring size of the use cases and > choosing productivity factor (hours per use case point) when > estimating based on use cases. The three other companies spent 587, > 829 and 943 hours respectively, showing that a heavier development > process and more emphasis on non-functional requirements may > increase effort by more than 100%. > > ... > > Section 3 describes the multiple-case study, i.e., the > requirements and the four development projects. > ... > > 4 Research Method > The research method was a multiple-case study with four development > projects developing > the same system in parallel. > Example: 16 case study sites of Clinical Leadership Programme, representing 8 regions. ------------------------------------------------------------------------ --------------------------------------- Here, study site = case. region = context. > Amultiple-case study evaluation of the RCN Clinical Leadership > Programme in England > > ... > The purpose ofthe > multiple-case study evaluation was to establish how key > stakeholders perceive the effectiveness and acceptability > ofthe CLP. > ... > > A multiple-case study approach was used for the > evaluation,incorporating a range ofmethods.Primarily, > these were undertaking qualitative interviews with key > stakeholders and an assessment ofdocumentary data. > ... > > A total of16 case study sites (two from each region) > were identified from the 80 English trusts taking part in > the CLP. This ensured a varied sample,utilising four > variables ofinterest;trust type,location,number of > directorates and annual patient turnover. > Example: 4 mixed-gender teams in two organizations: --------------------------------------------------- Unit is the team (see RQs). Organization is the context. > GENDERED PROCESSES IN SELF-MANAGING TEAMS: > A MULTIPLE CASE STUDY > Anne Marjukka Ollilainen > (her PhD thesis, Virginia Tech, Dept of Sociology) > > (ABSTRACT) > This study examines how gender as a socio-cultural construction > factors in the currently occurring > change from a bureaucratic work organization to more interactive > and team-based structures. > Informed by Joan AckerÕs theory of gendered organization, I > identify processes that produce and > reproduce gendered relationships of domination and subordination in > self-managing teams, despite > the premise that self-managing teams foster more egalitarian > workplace relations. In a multiple > case study, using in-depth interviews and participant observation, > I examine four currently > functioning, mixed-sex, self-managing teams in two service sector > organizations and one > manufacturing plant. The objective of the study is to uncover how > and in what ways gender is > present in teamwork and shapes various routine work processes. > The so-called Ògendered processesÓ I found to occur in the four > case-study teams include a gender > division of team tasks that required women to perform clerical work > even when teams were > supposed to implement cross-functional task sharing. Gendered > processes also took place through > interaction and team metaphors of ÔfamilyÕ and Ôfootball > teamÕ. I illustrate how the construction of > emotions in teamwork marginalized womenÕs contributions and how > women and men > consciously employed strategies to fit into expectations of gender- > appropriate behavior. Despite > these gender divisions, I suggest that one possible way for teams > to improve organizational gender > equality is that they emphasize non-hierarchical spatial > arrangements. Finally, although I found > gendered processes in all four teams, the ways in which gender > shaped teamwork varied according > to the organizational status position of a team. Also self- > management proved the most > comprehensive in teams that functioned at the higher organizational > levels. > Example: # cases = sample size (ie, each coalition is a case). ------------------------------------------------------------------------ -------- Unit of observation = multiple staffers of each coalition. > > Community health promotion relies heavily on coalitions to address > a multitude of public health issues. In spite of their widespread > use, there have been very few studies of coalitions at various > stages of coalition development. The purpose of this study was to > identify factors that facilitated or impeded coalition > effectiveness in the implementation stage of coalition development. > The research design was a multiple case study with cross-case > comparisons. Each of the 10 local North Carolina Project ASSIST > coalitions constituted a case. Data collection included: semi- > structured interviews, observation, document review, and surveys of > members and staff.