Conducting a Business Ethnography in Global Software Development projects of small German enterprises

Context

Studying work practices in the context of Global Software Development (GSD) projects entails multiple opportunities and challenges for the researchers. Understanding and tackling these challenges requires a careful and rigor application of research methods.

Objective

We want to contribute to the understanding of the challenges of studying GSD by reflecting on several obstacles we had to deal with when conducting ethnographically-informed research on offshoring in German small to medium-sized enterprises.

Method

The material for this paper is based on reflections and field notes from two research projects: an exploratory ethnographic field study, and a study that was framed as a Business Ethnography. For the analysis, we took a Grounded Theory-oriented coding and analysis approach in order to identify issues and challenges documented in our research notes.

Results

We introduce the concept of Business Ethnography and discuss our experiences of adapting and implementing this action research concept for our study. We identify and discuss three primary issues: understanding complex global work practices from a local perspective, adapting to changing interests of the participants, and dealing with micro-political frictions between the cooperating sites.

Conclusions

We identify common interests between the researchers and the companies as a challenge and chance for studies on offshoring. Building on our experiences from the field, we argue for an active conceptualization of struggles and conflicts in the field as well as for extending the role of the ethnographer to that of a learning mediator.     

Interactive observability in Ludics: The geometry of tests

Ludics [J.-Y. Girard, Locus solum, Math. Structures in Comput. Sci. 11 (2001) 301–506] is a recent proposal of analysis of interaction, developed by abstracting away from proof-theory. It provides an elegant, abstract setting in which interaction between agents (proofs/programs/processes) can be studied at a foundational level, together with a notion of equivalence from the point of view of the observer.

An agent should be seen as some kind of black box. An interactive observation on an agent is obtained by testing it against other agents.

In this paper we explore what can be observed interactively in this setting. In particular, we characterize the objects that can be observed in a single test: the primitive observables of the theory.

Our approach builds on an analysis of the geometrical properties of the agents, and highlights a deep interleaving between two partial orders underlying the combinatorial structures: the spatial one and the temporal one.     

Monofunctionalized β-cyclodextrins as sensor elements for the detection of small molecules

β-Cyclodextrins functionalized by different moieties that were tethered to a single 6-deoxyaminoglucose unit were investigated with respect to their suitability for sensor applications. Derivatizing the cyclodextrins with hydrophobic moieties like dipalmitoylglycerol and cholesterol allowed us to study packing density and orientation of the cyclodextrin tori at the air–water interface. From the pressure-area isotherms, it was concluded that the cyclodextrins are positioned towards the water subphase, with their molecular axis predominately parallel to the interface. By introducing a disulfide group, we managed to immobilize cyclodextrins on gold surfaces via self-assembly. MALDI mass spectrometry (MALDI MS) and XPS confirmed that the molecules are chemisorbed on the gold substrate displaying high surface coverage as determined by means of impedance spectroscopy. The inclusion of various charged guest molecules was monitored by changes in the charge transfer resistance of the redox couple [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻. The charge transfer resistance is sensitive to the surface potential, which leads to either repulsion or attraction of the redox active species.     

System requirements-OSS components: matching and mismatch resolution practices – an empirical study

Developing systems by integrating Open Source Software (OSS) is increasingly gaining importance in the software industry. Although the literature claims that this approach highly impacts Requirements Engineering (RE) practices, there is a lack of empirical evidence to demonstrate this statement. To explore and understand problems and challenges of current system requirement–OSS component matching and mismatches resolution practices in software development projects that integrate one or more OSS components into their software products. Semi-structured in-depth interviews with 25 respondents that have performed RE activities in software development projects that integrate OSS components in 25 different software development companies in Spain, Norway, Sweden, and Denmark. The study uncovers 15 observations regarding system requirements-OSS components matching and mismatch resolution practices used in industrial projects that integrate OSS components. The assessed projects focused mainly on pre-release stages of software applications that integrate OSS components in an opportunistic way. The results also provide details of a set of previously unexplored scenarios when solving system requirement–OSS component mismatches; and clarify some challenges and related problems. For instance, although licensing issues and the potential changes in OSS components by their corresponding communities and/or changes in system requirements have been greatly discussed in the RE literature as problems for OSS component integration, they did not appear to be relevant in our assessed projects. Instead, practitioners highlighted the problem of getting suitable OSS component documentation/information.     

Ranking-based evaluation of regression models

We suggest the use of ranking-based evaluation measures for regression models, as a complement to the commonly used residual-based evaluation. We argue that in some cases, such as the case study we present, ranking can be the main underlying goal in building a regression model, and ranking performance is the correct evaluation metric. However, even when ranking is not the contextually correct performance metric, the measures we explore still have significant advantages: They are robust against extreme outliers in the evaluation set; and they are interpretable. The two measures we consider correspond closely to non-parametric correlation coefficients commonly used in data analysis (Spearman's ρ and Kendall's τ); and they both have interesting graphical representations, which, similarly to ROC curves, offer useful various model performance views, in addition to a one-number summary in the area under the curve. An interesting extension which we explore is to evaluate models on their performance in “partially” ranking the data, which we argue can better represent the utility of the model in many cases. We illustrate our methods on a case study of evaluating IT Wallet size estimation models for IBM's customers. Saharon Rosset is Research Staff Member in the Data Analytics Research Group at IBM's T. J. Watson Research Center. He received his B.S. in Mathematics and M.Sc., in Statistics from Tel Aviv University in Israel, and his Ph.D. in Statistics from Stanford University in 2003. In his research, he aspires to develop practically useful predictive modeling methodologies and tools, and apply them to solve problems in business and scientific domains. Currently, his major projects include work on customer wallet estimation and analysis of genetic data. Claudia Perlich has received a M.Sc. in Computer Science from Colorado University at Boulder, a Diploma in Computer Science from Technische Universitaet in Darmstadt, and her Ph.D. in Information Systems from Stern School of Business, New York University. Her Ph.D. thesis concentrated on probability estimation in multi-relational domains that capture information of multiple entity types and relationships between them. Her dissertation was recognized as an additional winner of the International SAP Doctoral Support Award Competition. Claudia joined the Data Analytics Research group at IBM's T.J. Watson Research Center as a Research Staff Member in October 2004. Her research interests are in statistical machine learning for complex real-world domains and business applications. Bianca Zadrozny is currently an associate professor at the Computer Science Department of Federal Fluminense University in Brazil. Her research interests are in the areas of applied machine learning and data mining. She received her B.Sc. in Computer Engineering from the Pontifical Catholic University in Rio de Janeiro, Brazil, and her M.Sc. and Ph.D. in Computer Science from the University of California at San Diego. She has also worked as a research staff member in the data analytics research group at IBM T.J. Watson Research Center.     

Automatic feature selection for supervised learning in link prediction applications: a comparative study

For the last years, a considerable amount of attention has been devoted to the research about the link prediction (LP) problem in complex networks. This problem tries to predict the likelihood of an association between two not interconnected nodes in a network to appear in the future. One of the most important approaches to the LP problem is based on supervised machine learning (ML) techniques for classification. Although many works have presented promising results with this approach, choosing the set of features (variables) to train the classifiers is still a major challenge. In this article, we report on the effects of three different automatic variable selection strategies (Forward, Backward and Evolutionary) applied to the feature-based supervised learning approach in LP applications. The results of the experiments show that the use of these strategies does lead to better classification models than classifiers built with the complete set of variables. Such experiments were performed over three datasets (Microsoft Academic Network, Amazon and Flickr) that contained more than twenty different features each, including topological and domain-specific ones. We also describe the specification and implementation of the process used to support the experiments. It combines the use of the feature selection strategies, six different classification algorithms (SVM, K-NN, naïve Bayes, CART, random forest and multilayer perceptron) and three evaluation metrics (Precision, F-Measure and Area Under the Curve). Moreover, this process includes a novel ML voting committee inspired approach that suggests sets of features to represent data in LP applications. It mines the log of the experiments in order to identify sets of features frequently selected to produce classification models with high performance. The experiments showed interesting correlations between frequently selected features and datasets.     

Leadership Styles and Innovative Entrepreneurship: An International Study

This research attempts to empirically examine the relationship between leadership styles and innovative entrepreneurship through regression analysis, using a sample of 43 countries and data from Global Entrepreneurship Monitor and Global Leadership and Organizational Behavior Effectiveness. In light of institutional approaches and specifically based on the normative dimension, the main findings of the study indicate that participative leadership and higher education represent the strongest explanatory factor in the variance of the current rates of innovative entrepreneurship. This study has contributions for both researchers and policymakers on new firm creation (entrepreneurship) and on the generation of innovation within organizations (intrapreneurship).     

Cognitive load and science text comprehension: Effects of drawing and mentally imagining text content

One hundred and eleven 10th graders read an expository science text on the dipole character of water molecules (ca. 1600 words). Reading instruction was varied according to a 2 × 2 experimental design with factors ‘drawing pictures of text content on paper’ (yes, no) and ‘mentally imagining text content while reading’ (yes, no). The results indicate that drawing pictures, mediated through increased cognitive load, decreased text comprehension and, thus, learning (d = −0.37), whereas mental imagery, although decreasing cognitive load, increased comprehension only when students did not have to draw pictures simultaneously (d = 0.72). No evidence was found that the effects were moderated by domain-specific prior knowledge, verbal ability, or spatial ability. The results are in line with cognitive theories of multimedia learning, self-regulated learning, and mental imagery as well as conceptions of science learning that focus on promoting mental model construction by actively visualizing the content to be learned. Constructing mental images seems to reduce cognitive load and to increase comprehension and learning outcome when the mental visualization processes are not disturbed by externally drawing pictures on paper, whereas drawing pictures seems to increase cognitive load resulting in reduced comprehension and learning outcome.     

Discovering anomalous frequent patterns from partially ordered event logs

Conformance checking allows organizations to compare process executions recorded by the IT system against a process model representing the normative behavior. Most of the existing techniques, however, are only able to pinpoint where individual process executions deviate from the normative behavior, without considering neither possible correlations among occurred deviations nor their frequency. Moreover, the actual control-flow of the process is not taken into account in the analysis. Neglecting possible parallelisms among process activities can lead to inaccurate diagnostics; it also poses some challenges in interpreting the results, since deviations occurring in parallel behaviors are often instantiated in different sequential behaviors in different traces. In this work, we present an approach to extract anomalous frequent patterns from historical logging data. The extracted patterns can exhibit parallel behaviors and correlate recurrent deviations that have occurred in possibly different portions of the process, thus providing analysts with a valuable aid for investigating nonconforming behaviors. Our approach has been implemented as a plug-in of the ESub tool and evaluated using both synthetic and real-life logs.     

Proteomic analysis to improve adequacy of hemo- and peritoneal dialysis: Removal of small and high molecular weight proteins with high- and low-flux filters or a peritoneal membrane

The reduction of morbidity and mortality in patients undergoing hemo- or peritoneal dialysis is strongly related to an efficient and selective clearance of uremic toxins. We used proteomics methods to analyze and further characterize the dialytic removal of still undefined middle and high molecular weigh proteins as a basis for further improvement of dialysis assessment. Dialysates from 26 hemodialysis patients treated with different types of low- (F6HPS?) and high-flux (FX80?, APS650?, FX60?) filters as well as peritoneal fluids from 10 continuous ambulatory peritoneal dialysis (CAPD) patients were analyzed by SELDI-TOF and 2-DE. The protein patterns showed selective differences in the proteins cleared depending on the dialysis method used and the filter membrane. While SELDI analyses of dialysates from the F6HPS revealed almost no protein clearance, high-flux filters and CAPD dialysates showed protein release of different molecular weight ranges. Furthermore, 2-DE and MS analysis identified 48 different proteins from the dialysate of high-flux filters and 21 from peritoneal dialysis fluids. In F6HPS dialysates, however, only few proteins could be identified.