Many students find that learning statistics is difficult, especially novice learners first taking statistical classes. Instructors also struggle to teach statistics due to many challenges, such as students' misconceptions, lack of math skills, negative attitudes, and deficient instructional methods. A new pedagogical approach called Decision-Based Learning (DBL) has been developed to help instructors and students overcome these challenges. A group of Peruvian university professors in 2018 and 2019 were trained to use DBL. However, limited studies have explored DBL instructors' and students' experiences using DBL in Peru. The aim of this qualitative research study is to deeply understand the experiences of one Peruvian statistics instructor and her students using DBL pedagogy and software. This study used a qualitative Gadamerian-Hermeneutic methodology approach. Data were collected from seven participants' interviews (six students and one statistics instructor), the researcher's field notes, and classroom artifacts. Findings identify several specific benefits and challenges participants faced when implementing DBL. Benefits include growth of students' confidence in solving statistical problems and instructor satisfaction, and challenges includes time-consuming instructor initial preparation and students' learning transfer issues. This study concluded that DBL is beneficial for students and instructors alike. Based on the results, some recommendations are offered to help instructors to overcome the challenges faced when implementing DBL.

This study evaluated whether significant differences in statistical reasoning abilities exist for completers of short online instructional videos and formative quizzes for students in undergraduate introductory statistics courses. Data for the study were gathered during the Fall 2013 semester at a community college in Northeast Tennessee. Computer-based pedagogical tools can promote improved conceptual reasoning ability (Trumpower & Sarwar, 2010; Van der Merwe, 2012). Additionally, prior research demonstrated a significant relationship between formative quiz access and student achievement (Stull, Majerich, Bernacki, Varnum, & Ducette, 2011; Wilson, Boyd, Chen, & Jamal, 2011), as well as multimedia object access and student achievement (Bliwise, 2005; Miller, 2013). Four research questions were used to guide the study. A series of analysis of variance (ANOVA) statistical procedures was used to analyze the data. Findings indicated no significant differences in statistical reasoning abilities between students who were provided access to supplemental online instructional videos and formative quizzes and students who were not provided access. Moreover, statistical reasoning abilities did not differ significantly based upon number of quizzes successfully completed, average number of quiz attempts, or number of videos accessed.

Made available in DSpace on 2016-04-27T16:57:08Z (GMT). No. of bitstreams: 1 Diva Valerio Novaes.pdf: 2823527 bytes, checksum: 17ea15131520d732f9fb8c60ce4d5769 (MD5) Previous issue date: 2011-08-10 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This study sought to investigate conceptions about objects of Descriptive Statistics related to teaching or to specific content that 1st- to 12th-grade Brazilian teachers employ when organizing and managing teaching sequences focused on this theme. To this end, the procedures adopted in the classroom by teachers attending a continued education program were observed. The investigation was carried out in response to the finding, by previous studies conducted in this field, of barriers to the learning of notions of Statistics. Selection of the objects of study was based on identification of their mutual relationships in analogy to those established within a stable ecosystem herein assumed in analogy to the biological notion of the food chain. The research was part of the project entitled Teaching and Learning Process Involving Probabilistic and Statistical Thinking (PEA-ESTAT), funded by the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP). Based on the case study method, the investigation sought to address the following questions: What ideas can be identified when 1st- to 12th-grade teachers apply their statistical knowledge of variation to solve problems and prepare their lessons on this topic? How can this knowledge be modeled with the aid of the Theory of Conceptions so as to establish parameters that contribute to overcoming or minimizing barriers and difficulties associated to the learning of statistical content, as identified in studies in this area? The ck¢ model was adopted to describe and explain the conceptions expressed by the teachers participating in this study. This model provided a theoretical framework that allowed plausible explanations to be inferred for cognitive procedures that generate many of the problems identified in other studies, with regard to eliciting the knowledge required to conduct an exploratory data analysis that leads to correct comprehension of the concept of variability. Sixteen consistently interconnected teaching- and statistics-related conceptions were identified, which provided answers to the questions posed in this investigation / Esta pesquisa teve por objetivo analisar concepções sobre objetos da Estatística Descritiva, tanto didáticas quanto específicas de conteúdo, mobilizadas por professores da Educação Básica quando organizam e fazem a gestão de sequências didáticas nesse tema e, para tanto, observamos os procedimentos adotados em sala de aula por professores em formação continuada. A pesquisa teve origem na constatação da existência de entraves à aprendizagem das noções estatísticas identificados em outros estudos na área. A escolha dos objetos de estudo foi feita a partir da identificação das relações estabelecidas entre eles em analogia com as que se estabelecem em um ecossistema estável, por sua vez assumido em analogia à noção biológica de cadeia alimentar. A pesquisa está inserida no projeto Processo de Ensino e Aprendizagem Envolvendo Pensamento Estatístico e Probabilístico (PEA-ESTAT), financiado pela Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP). A metodologia utilizada foi o estudo de caso, com o qual buscamos responder às seguintes questões: Quais concepções podem ser identificadas quando professores da Educação Básica mobilizam seus conhecimentos estatísticos sobre variação ao resolverem problemas e prepararem suas aulas sobre esse tema? Como esses conhecimentos podem ser modelados com auxílio da Teoria das Concepções, de modo a se estabelecerem parâmetros que contribuam para a superação ou minimização de entraves e dificuldades de aprendizagem desses conteúdos estatísticos, já identificados em pesquisas na área? Utilizamos o modelo ck¢ para descrever e explicar as concepções manifestadas pelos professores que foram sujeitos nesta pesquisa. Tal modelo proporcionou um quadro teórico que permitiu inferir, a partir dos dados coletados, explicações plausíveis para procedimentos cognitivos que geram diversas das dificuldades já identificadas em outros estudos, no que se refere a mobilizar os conhecimentos necessários para realizar uma análise exploratória de dados que conduza à correta apreensão do conceito de variabilidade. Identificaram-se 16 concepções didáticas e estatísticas que sempre funcionam interrelacionadas, o que nos permitiu responder às questões propostas nesta pesquisa

The crucial role of representation in mathematical and statistical modeling and problem solving as evident in learners' arrangement and representation of statistical data were investigated with focus points data arrangement, data representation and statistical thinking levels. The representation tasks required learners to arrange and represent data through modeling, focusing on spontaneous representations. Successful transnumeration determine the ultimate success of a representation and the ability to organise data is regarded as critical. Arrangement types increased in sophistication with increased grade level and the hierarchical nature of arrangement types became apparent when regarded in the context of an adapted SOLO Taxonomy framework. A higher level arrangement strategy pointed to a higher SOLO level of statistical thinking. Learners in the two tasks produced a rich variety of representations which included idiosyncratic, unsophisticated responses as well as standard statistical representations. The context of the two tasks, the quantitative versus qualitative nature of the data in the tasks, and the statistical tools or representational skills learners have at their disposal, played an important role in their representations. Well-planned data handling activities develop representational and higher order thinking skills. The variety of responses and different response levels elicited in the two tasks indicate that the nature of the tasks rather than the size of the data set play a conclusive role in data tasks. Multiple representations by an individual were an indication of successful modeling, are effective in problem solving and are associated with good performance. The SOLO model which incorporates a structural approach as well as a multimodal component proved valuable in the analysis of responses. Using this model with accompanying acknowledgement of different problem solving paths and the contribution of ikonic support in the concrete symbolic mode, promote the in-depth analysis of responses. This study contributes to the research in the field of data representation and statistical thinking. The analysis and results led to an integrated picture of Grade 4-7 learners' representation of statistical data and of the statistical thinking levels evident in their representations. / Educational Studies / D. Ed. (Didactics)

Made available in DSpace on 2016-04-27T16:57:25Z (GMT). No. of bitstreams: 1 Marcelo Massahiti Yamauti.pdf: 2215151 bytes, checksum: c412cbeecb2abae05c4b982bf1270e02 (MD5) Previous issue date: 2013-05-14 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This research is inserted in the context of an Administration course and its purpose is to analyze how the Statistics textbooks for such courses organized the proposed activities referring to the study of the Simple Linear Regression theme, verifying if these contributed to the development of the Statistical Thinking. With the intent of reaching such purpose, six textbooks were analyzed, considering the proposed exercises in the referring chapter to the theme in hand. The theoretical foundation on which the analysis was based is the Didactical Anthropological Theory, proposed by Yves Chevallard in the sight of the praxeological organization. Besides this, the founding defined as Statistical Thinking proposed by many authors of the Statistical Didactical area was also used. Based on these theories, the analyses were held with the intent of answering the two problem questions. Which praxeological organizations (Statistics/Mathematics) the Statistics textbooks for Administration courses show the Simple Linear Regression content? Which characteristics of the Statistical Thinking are met in these identified organizations? Thus, a qualitative research was held with a documental focus. The results of the textbook analyses pointed out for the fact that the professor needs to be more reasonable in the choice of a proposed activity, if the purpose was the development of the student s Statistical Thinking / Esta pesquisa insere-se no contexto de um curso de Administração e teve por objetivo analisar como os livros-texto de Estatística para tais cursos organizaram as atividades propostas, referentes ao estudo do tema Regressão Linear Simples, verificando se estas contribuem para o desenvolvimento do Pensamento Estatístico. Com o intuito de atingir tal objetivo, foram analisados seis livros-texto, considerando os exercícios propostos no capítulo referente ao tema em questão. O quadro teórico que fundamentou as análises foi a Teoria Antropológica do Didático, proposta por Yves Chevallard com o olhar da organização praxeológica. Além deste, buscou-se também a fundamentação no que se define como Pensamento Estatístico, proposto por vários autores da área da Didática da Estatística. Apoiado nessas teorias, as análises foram realizadas com o intuito de responder as duas questões de pesquisa: Quais organizações praxeológicas (Estatística/Matemática) os livros-texto de Estatística para cursos de Administração apresentam em relação ao conteúdo Regressão Linear Simples? Quais características do Pensamento Estatístico são contempladas nessas organizações identificadas? Para isto, foi desenvolvida uma pesquisa qualitativa com enfoque documental. Os resultados das análises dos livrostexto apontaram para o fato de que o professor precisa ser mais criterioso na escolha de uma atividade proposta, se o objetivo a ser alcançado for o desenvolvimento do Pensamento Estatístico dos alunos

Made available in DSpace on 2016-04-27T16:57:43Z (GMT). No. of bitstreams: 1 dissertacao_tula_rocha.pdf: 569614 bytes, checksum: 099181350e8629e18d86379838179bdd (MD5) Previous issue date: 2006-06-01 / The present research has for purpose to investigate the conceptions of teachers of basic teaching about statistical mind. Bellow the referential of mind s components of Gal (2002), the levels of statistical literacy of Shamos (1995) and the measurements of pathern PPDAC (Problem, scheme, data, analyse and conclusion) of Wild e Pfannkuch (1999), is accomplished the investigation for means of diagnostic instrument and analyse of didactics books of mathematic. This job introduce the study about the statistical mind, explanations for nim, components and abilities necessaries to statistical literacy. In the following, we analysed two collections of didactics books bellow the viewpoint of Praxeology Organization (CHEVALLARD, 1996), with how to identify the tasks, technic and theoritical-technological discourse, privileged for them. The study of didactic books, allowed identify, of current teaching, the approach tecnicist of statisti of the fundamental teaching. Toward investigate the teacher s conception, such as the influence of the didactic book in her lecturing skill, applied the questionnaire in mathematic teachers of the basic and intermediate teaching of Belo Horizonte s schools. The results was submitted at analysis helped of software C.H.I.C. (Classification, hierarchic, implicative and cohesitive), with how to associate the teachers s responses. The analyse that allowed to conclued that the teachers develop statistics abilities propitious to literacy on cultural level, instead of appropriate functional, of this shool segment. Maybe, influenced by the didactics books that support seeing through statistc of tecnicist, emphasising the use of tabulares record and graphics, beyond the algoritimic interpretation of conception to arithmetic s mean / A presente pesquisa tem por objetivo investigar as concepções de professores do Ensino Fundamental sobre o Pensamento Estatístico. Sob o referencial dos Componentes do Pensamento de Gal (2002), dos níveis de letramento estatístico de Shamos (1995) e das dimensões do Modelo PPDAC (Problema, Plano, Dados, Análise e Conclusão) de Wild e Pffannkuch (1999) é realizada a investigação por meio de um instrumento diagnóstico e da análise de livros didáticos de Matemática. Iniciamos este trabalho com um estudo sobre o Pensamento Estatístico, definições a ele atribuídas, componentes e habilidades necessários ao letramento estatístico. Em seguida, analisamos duas coleções de livros didáticos a luz da Organização Praxeológica (Chevallard, 1996), de modo a identificar as tarefas, técnicas e discurso teórico-tecnológico por eles privilegiados. O estudo de livros didáticos permitiu identificar, no ensino atual, a abordagem tecnicista da Estatística no Ensino Fundamental. Para investigar a concepção dos professores, bem como a influência do livro didático em sua prática docente, aplicamos um questionário para 20 professores de Matemática do Ensino Fundamental e Médio de escolas de Belo Horizonte. Os resultados foram submetidos à análise auxiliada pelo software C.H.I.C. (Classificação Hierárquica, Implicativa e Coesitiva). A análise permitiu inferir habilidades estatísticas propícias ao letramento no nível «cultural», ao invés do «funcional» adequado a esse segmento escolar. Possivelmente influenciado pelos livros didáticos que favorecem uma visão tecnicista da Estatística, priorizando o uso de registros tabulares e gráficos, além da interpretação algorítmica do conceito de média aritmética

Made available in DSpace on 2016-04-27T16:57:53Z (GMT). No. of bitstreams: 1 dissertacao_michele_medici.pdf: 6827447 bytes, checksum: 0b99bfd53a4ccff480f5166c93263315 (MD5) Previous issue date: 2007-05-17 / Either during our every day life or in studies or scientific researches, we evidenced the necessity of exploring tabular and graphic representations. The report from the 4th INAF stands out, among other aspects, that only 23% of the Brazilian population shows some familiarity with this kind of representations. As we believe that the sooner we begin to explore them, the better. The aim of this essay was to conceive a didactic sequence on an experimental focus in order to introduce statistics to 5th graders (nowadays belonging to the 6th year) from Ensino Fundamental (Secondary School). We are not only looking for didactic conditions that would support student s autonomous evolution on solving problems of organization, representation and interpretation of a set of data, as well as the didactic sequence which the teacher would use in order to promote the construction of the statistical thinking. Thus, we investigate the way the students interact with situations proposed by the teachers, the students´ former knowledge, hypothesis formulated by them and the way they deal with constructed knowledge. For such analysis, we made use of the surmises of Didactic Engineering and we concluded that the classes must be fulfilled with collective debates and that the students have to be in small groups and every step must be built by them and they are responsible for their research. We are able to notice that the elements for the construction of the statistical thinking could be gradually composed by the students and the representations were most of the time, badly organized and / or with incorrect or missing information. The debates led to homogeneity of the milieux which turned into the students´ learning. We raised a series of questions to be explored with them during their following school year / No nosso dia-a-dia ou em estudos e pesquisas científicas constatamos a necessidade de explorar as representações tabulares e gráficas. O relatório do 4º INAF ressalta, entre outros aspectos, que apenas 23% da população brasileira demonstra certa familiaridade com essas representações. Por acreditar que devemos começar cedo a explorá-las, o objetivo desta dissertação foi conceber uma seqüência didática, em um enfoque experimental, para introduzir estatística aos alunos da 5ª série (hoje 6º ano) do Ensino Fundamental. Buscamos não apenas as condições didáticas que favoreçam a evolução autônoma do aluno na resolução de problemas de organização, representação e interpretação de um conjunto de dados, mas também a seqüência didática que o professor possa utilizar, visando favorecer a construção do pensamento estatístico. Assim, investigamos a maneira como o aluno interage com as situações propostas pelo professor, os conhecimentos preliminares que os alunos já possuem, as hipóteses elaboradas por eles e a forma como mobilizam os conhecimentos construídos. Para tais análises, utilizamos os pressupostos da Engenharia Didática e concluímos que as aulas devem ser permeadas por debates coletivos e em pequenos grupos de trabalho e que todas as etapas devem ser construídas pelos alunos, responsáveis pela sua pesquisa. Pudemos verificar que os elementos para a construção do pensamento estatístico puderam ser compostos gradativamente pelos alunos e que as representações foram muitas vezes pouco organizadas e / ou com informações inexatas ou faltantes. Os debates proporcionaram uma homogeneização dos milieux, o que pôde levar à aprendizagem dos alunos. Levantamos uma série de questões para serem exploradas com esses alunos no ano seguinte de escolaridade

The crucial role of representation in mathematical and statistical modeling and problem solving as evident in learners' arrangement and representation of statistical data were investigated with focus points data arrangement, data representation and statistical thinking levels. The representation tasks required learners to arrange and represent data through modeling, focusing on spontaneous representations. Successful transnumeration determine the ultimate success of a representation and the ability to organise data is regarded as critical. Arrangement types increased in sophistication with increased grade level and the hierarchical nature of arrangement types became apparent when regarded in the context of an adapted SOLO Taxonomy framework. A higher level arrangement strategy pointed to a higher SOLO level of statistical thinking. Learners in the two tasks produced a rich variety of representations which included idiosyncratic, unsophisticated responses as well as standard statistical representations. The context of the two tasks, the quantitative versus qualitative nature of the data in the tasks, and the statistical tools or representational skills learners have at their disposal, played an important role in their representations. Well-planned data handling activities develop representational and higher order thinking skills. The variety of responses and different response levels elicited in the two tasks indicate that the nature of the tasks rather than the size of the data set play a conclusive role in data tasks. Multiple representations by an individual were an indication of successful modeling, are effective in problem solving and are associated with good performance. The SOLO model which incorporates a structural approach as well as a multimodal component proved valuable in the analysis of responses. Using this model with accompanying acknowledgement of different problem solving paths and the contribution of ikonic support in the concrete symbolic mode, promote the in-depth analysis of responses. This study contributes to the research in the field of data representation and statistical thinking. The analysis and results led to an integrated picture of Grade 4-7 learners' representation of statistical data and of the statistical thinking levels evident in their representations. / Educational Studies / D. Ed. (Didactics)

Submitted by Livia Mello (liviacmello@yahoo.com.br) on 2016-09-28T19:01:15Z No. of bitstreams: 1 DissNGC.pdf: 2769660 bytes, checksum: fed4a63a83c3d31c36b783f9c91ef798 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-10-04T18:57:37Z (GMT) No. of bitstreams: 1 DissNGC.pdf: 2769660 bytes, checksum: fed4a63a83c3d31c36b783f9c91ef798 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-10-04T18:57:44Z (GMT) No. of bitstreams: 1 DissNGC.pdf: 2769660 bytes, checksum: fed4a63a83c3d31c36b783f9c91ef798 (MD5) / Made available in DSpace on 2016-10-04T18:57:51Z (GMT). No. of bitstreams: 1 DissNGC.pdf: 2769660 bytes, checksum: fed4a63a83c3d31c36b783f9c91ef798 (MD5) Previous issue date: 2015-10-16 / Não recebi financiamento / With more advanced marketing relations, various changes are taking place in regard to competitive factor markets, among which, the importance of quality processes and goods supplied to the society, stand out. Therefore, diverse programs and methods have been developed to monitor indicators of production processes and the characteristics of roduct quality, many of which are supported by statistical concepts. This process encourages understanding and the use of Statistical Thinking (ST), Statistical Engineering (SE), and Statistical Techniques (STe) in businesses, highlighting their importance to quality management and the efficiency of organizational processes. This thesis’s aim was to analyze the use of ST and STe in a sample of industries in the sector of serial capital goods located in the state of São Paulo, Brazil. For that, an exploratory-descriptive survey was conducted with 42 medium and large companies from different segments of the sector. Data were statistically analyzed using descriptive and multivariate methods (cluster analysis, association and correlation). These analyses revealed there is considerable space to diffuse Statistical Thinking and Statistical Techniques in these companies, as well as revealed a lack of training in statistical principles and techniques, the main obstacles to the systematic application of these techniques, and the lack of support on the part of top management. There is evidence that the successful implementation of Total Quality Management (TQM) and Total Productive Maintenance in the industries of sector significantly influences the diffusion of ST and the application of STe; while the use of Six Sigma program especially impacts the application of more complex STe; and ST has the potential to promote the application of STe. / Com a evolução das relações mercadológicas, diversas mudanças vêm ocorrendo a respeito dos fatores competitivos do mercado, dentre os quais se destaca a importância da qualidade dos processos e produtos disponibilizados à sociedade. Dessa maneira, diversos programas e métodos têm sido desenvolvidos com o objetivo de acompanhar os indicadores dos processos produtivos e monitorar as características da qualidade do produto, utilizando para isso técnicas, muitas delas suportadas por conceitos estatísticos, o que, consequentemente, estimula a compreensão e uso do Pensamento Estatístico (PE), da Engenharia Estatística (EE) e das Técnicas Estatísticas (TE) nas empresas e destaca sua importância para a gestão da qualidade e eficiência de processos organizacionais. Essa dissertação tem por objetivo analisar o uso do PE e de TE em uma amostra de empresas industriais do setor de bens de capital seriados, localizadas no estado de São Paulo. Para isso, foi conduzida uma survey exploratória-descritiva em 42 empresas de médio e grande porte de diferentes segmentos do setor. Os dados obtidos foram analisados estatisticamente por métodos descritivos e multivariados (análise de cluster, associação e correlação). Tais análises revelaram a existência de um grande espaço para difusão do Pensamento Estatístico e das Técnicas Estatísticas nessas empresas, bem como a ausência de treinamentos em princípios e técnicas estatísticas, os principais obstáculos para a aplicação sistemática dessas técnicas e a falta de suporte da alta administração. Há evidências de que a implantação bem-sucedida dos programas TQM (Total Quality Management) e TPM (Total Productive Maintenance) nas empresas o setor exerceu significativa influência sobre a difusão do PE e aplicação das TE; de que o Programa Seis Sigma impacta especialmente na aplicação das TE de maior complexidade; e de que o uso do PE tem potencial para promover maior aplicação das TE.

Made available in DSpace on 2016-04-27T16:58:16Z (GMT). No. of bitstreams: 1 Claudia Borim da Silva.pdf: 1181639 bytes, checksum: a1c349e009d97a59b67411221aef0c9a (MD5) Previous issue date: 2007-05-30 / Due to student difficulty with understanding standard deviation, this work aimed to identify the reasoning about variation and variability in all parts of the investigation cycle of the statistical thinking. Nine middle and high school Mathematics teachers and two mathematics students of University of São Paulo participated in an action research, 3 hr meetings, lasting in total for 48 hrs. The contents were simple and grouping data frequency distribution, graphics, center and spread measures. The reasoning levels were classified using the general model developed by Garfield (2002). The teachers showed no variaton reasoning during the first week, except for a teacher with idiosyncratic reasoning. During the action research sensibility phase and planning of investigative cycle phase, the teachers developed variability reasoning naturally, but not about variation. However, this experience promoted an upgrade of teachers statistical thinking, that used three (between four) dimensions created by Wild e Pfannkuch (1999). Nevertheless, the statistical thinking upgrade did not implicate a gain in variation reasoning level, observed during the data analysis phase. To compare three discret variable frequency distribution were done using the perception of mode, minimum and maximum values and minimum frequency and use of the distribution chunk with range was organized with existence of the frequency in all groups, understood like verbal until procedural reasoning, respectively. The center measures discussion showed the misconception of mean, which was understood as the mode, and this inhibited necessity perception of a spread measure. The use of correct mean of arithmetic mean induced the teachers use complement measures as the mode and minimum and maximum values, but not the standard deviation. The mean fo standard deviation was predominantely a measure of number of differents observations, signal of homogeneous sample, as many Mathematics textbooks introduced the concept of variation. The comprehension of one standard deviation interval towards mean didn´t develop naturally and the teachers who understood this mean of standard deviation had difficulty to understand what was in the interval, which suposed to develop this integrated reasoning process with the educational softwares created for this intention. In conclusion, the term more variation can cause wastly differing results due to personal interpretation of the phrase ´more variation´ and idiosyncratic reasoning process involved in analysing complex mathematical data: more variation between frequency in only the variable category or variable value in comparing frequency distributions and more variation between sample different observations, both without use of variation from mean / Devido à dificuldade encontrada por alunos de graduação para a compreensão do desvio padrão, este trabalho teve como objetivo verificar o raciocínio sobre variação e variabilidade nas etapas do ciclo investigativo do pensamento estatístico. Foram participantes da pesquisa nove professores de Matemática da escola básica e dois alunos de Matemática da Universidade de São Paulo. O trabalho seguiu os pressupostos de uma pesquisa-ação e a fase de implementação teve duração de quarenta e oito horas, divididas em dezesseis encontros de três horas cada. Foram discutidos os conteúdos estatísticos: distribuição de freqüência simples e com dados agrupados, representações gráficas, medidas de tendência central e dispersão. Os níveis de raciocínio sobre variação foram classificados de acordo com o modelo proposto por Garfield (2002). O diagnóstico identificou a ausência de raciocínio sobre variação, exceção feita a um professor que apresentava raciocínio idiossincrático. Durante a fase de sensibilização da pesquisa-ação e planejamento do ciclo investigativo, os professores apresentaram naturalmente o raciocínio sobre variabilidade, mas não sobre variação. Entretanto, a experiência com a elaboração de uma pesquisa, desde a definição dos objetivos até a coleta e montagem do banco de dados permitiu um avanço no desenvolvimento do pensamento estatístico dos professores, que já transitavam em três das quatro dimensões de sua estrutura elaborada por Wild e Pfannkuch (1999). Não obstante, o desenvolvimento do pensamento estatístico não implicou diretamente em um nível mais avançado do raciocínio de variação, observado durante a fase de análise dos resultados da pesquisa. Para a comparação de três distribuições de freqüências simples de variável discreta foram utilizadas a percepção da moda, a observação dos valores máximo e mínimo e da menor freqüência e a elaboração de um intervalo de variação composto pelos valores da variável que tinham freqüência nas três distribuições, conjuntamente, que foram categorizados como raciocínio verbal de variação até raciocínio de procedimento, respectivamente. A discussão sobre as medidas de tendência central permitiu observar a interpretação equivocada de média como maioria, que se refere à moda, que foi um fator impeditivo para a percepção da necessidade de uma medida de variação. A utilização do correto significado de média motivou os professores a utilizarem medidas complementares como a moda e os valores máximo e mínimo, mas não o desvio padrão. O significado atribuído ao desvio padrão foi, predominantemente, uma medida da variação entre as observações indicando homogeneidade da amostra, aspecto reforçado pelos livros didáticos de Matemática do ensino médio e categorizado como raciocínio verbal de variação. A composição do intervalo de um desvio padrão da média não surgiu naturalmente e mesmo os participantes que compreenderam esta interpretação do desvio padrão, apresentaram dificuldade para identificar o que tinha no intervalo. Acredita-se que o desenvolvimento de aplicativos computacionais para trabalhar o conceito de intervalo em torno da média possa auxiliar na aquisição deste raciocínio, considerado um raciocínio completo de variação. Conclui-se que a linguagem maior variação pode induzir dois diferentes raciocínios idiossincráticos: a maior variação das freqüências em alguma categoria ou valor da variável de uma distribuição de freqüências e a maior variação de observações diferentes na amostra, ambas não relacionadas com a medida de tendência central