For the last couple of decades consumer decision-making has been of increasing interest for retail as well as for consumer behaviour research. Food shopping constitutes a unique type of shopping behaviour. In comparison to other types of shopping, food is essential to life, and not often are there as many choices to be made in a short period of time as when shopping groceries. The purpose of this dissertation was to advance the knowledge of what influences consumers’ choices of grocery stores. More specifically, the main focus has been on how different situations (e.g., type of shopping) influence choices of grocery stores. Five papers, which build on three surveys on how consumers choose grocery stores in Sweden, are included in this dissertation. In the first paper a comprehensive set of ten aggregated attributes that determine store choices were developed. The second paper brought forward five consumer segments (Planning Suburbans, Social Shoppers, Pedestrians, City Dwellers, and Flexibles) based on where and how they shop. In the third paper it was shown that accessibility attributes (e.g., accessibility by car, availability) and attractiveness attributes (e.g., price, service) have different impacts on satisfaction, depending on consumer characteristics and shopping behaviour in supermarkets compared to convenience stores. In the fourth paper the result showed that satisfaction is affected by type of grocery shopping (major versus fill-in shopping) in conjunction with time pressure and which store attributes that are important for satisfaction. It was also shown that the effect of time pressure and type of shopping on satisfaction varied in different consumer segments. In the final paper it was shown that a store has to be more attractive in terms of attributes for a consumer to switch from the grocery store they usually patronage, even if the new store is situated right beside or closer than the consumer’s regular grocery store. The view of a “good location” is further developed in this dissertation, arguing that consumers’ mental distance to a store – their cognitive proximity – is much more important than the physical place of the store. In sum, this dissertation revealed that the situation is more important than previous research has shown. Depending on the situation, consumers will face different outcomes (different stores) and value different store attributes. Hence, stores need to manage different store attributes depending on which consumer groups the stores want to attract and what situation the consumers are facing. Therefore, consumers’ choices of grocery stores are situation-based choices. / I ett par decennier har intresset för konsumenters beslutsfattande ökat för både detaljhandeln och forskningen kring konsumentbeteende. Matinköp utgör en unik typ av köp-beteende då det i jämförelse med andra typer av handlande är livsnödvändiga samt att det sällan finns så många val som ska göras under kort tid som vid matinköp. Syftet med denna avhandling är att främja kunskap om vad som påverkar konsumenternas val av livsmedelsbutik. Mer specifikt har fokus varit på hur olika situationer (t.ex. typ av handlande) påverkar valet av butik. Fem artiklar, som bygger på tre olika undersökningar om hur konsumenter väljer livsmedelsbutiker i Sverige ingår i denna avhandling. I den första artikeln utvecklades en omfattande uppsättning av tio aggregerade attribut (baserade på 34 attribut) som bestämmer konsumenters val av livsmedelsbutiker. I den andra artikeln presenterades fem konsumentsegment (Planerande förortsbor, Sociala shoppare, Fotgängare, Stadsbor och Flexibla) som baserades på var och hur de handlar. Den tredje artikeln visade att tillgänglighetsattribut (t.ex. tillgängligheten med bil och öppettider) och attraktivitetsattribut (t.ex. pris och service) har olika effekter på konsumenters nöjdhet. Denna nöjdhet varierade även beroende på konsumentens bakgrundsfaktorer samt huruvida konsumenten handlade i stormarknader eller i närbutiker. I den fjärde artikeln visade resultaten att nöjdhet påverkas av typ av matinköp (storhandlande kontra kompletteringshandlande) i samband med tidspress och de attribut som är viktiga för konsumenternas nöjdhet med butiken. Det visade sig även att effekterna av tidspress och typ av handlande på konsumenternas nöjdhet med butiker varierade i olika konsumentgrupper. Det femte konferenspapperet visade att en butik måste vara mer attraktiv när det gäller attribut för att konsumenter skall byta från den livsmedelsbutik som de brukar handla i, även om den nya butiken skulle öppna precis bredvid eller närmre än den vanliga livsmedelsbutiken. Synen på vad som är ett ”bra läge” utvecklas därför ytterligare i denna avhandling, med argumentet att konsumenternas mentala avstånd till en butik - deras kognitiva närhet - är mycket viktigare än den fysiska platsen för butiken. Sammanfattningsvis visade denna avhandling att effekten av olika situationer är viktigare än vad tidigare forskning har visat. Beroende på situation kommer konsumenter att möta olika utfall (välja olika butiker) och de kommer även att värdera olika butikers attribut olika. Således behöver butiken hantera olika butiksattribut beroende på vilken konsumentgrupp butiken vill attrahera och vilken situation de konsumenterna står inför. Därför kan val av livsmedelsbutiker ses som situationsbaserade val.

This thesis contributes to recent theorizations in economic geography on the effects of proximity on firm competitiveness. One of the great challenge in the contemporary economy is for firms to remain competitive. Their innovative ability is highly dependent on the knowledge they possess and their ability to acquire new knowledge. It is argued that a relational proximity between individuals reduces uncertainty and offers a joint platform for communication and learning. Therefore, does this thesis apply a micro-perspective in which the labor force and the knowledge composition within plants is examined. The aim is pursued by exploring the interrelationship between different types of proximity in the labor force and plant performance. The proximity dimensions under scrutiny are; the cognitive-, the organizational-, and the geographical proximity dimension. The three empirical papers in the thesis are based on longitudinal micro-data from the database ASTRID. The database connects detailed socio-economic data of individuals to features of plants and firms in the entire Swedish economy. The empirical findings suggest that the different types of proximities are interdependent with regard to learning in firms. The interdependence is manifested through the variable impact on plant performance that a given distance in one dimension has, depending on what other type of proximity is accounted for at the plant. It is further found that the proximity dimensions have conditional effects on learning and innovation in firms. The empirical findings also indicate that the circumstances under which learning and knowledge application take place, vary between capital-intensive and labor-intensive sectors. Moreover, it is found that relatedness in the cognitive dimension is not unambiguously positive for interactive learning and innovativeness. Similarity in one dimension and unrelatedness in the cognitive dimension, has a significantly stronger impact on interactive learning than simply having relatedness in the cognitive dimension. It therefore seems as if the combined distance of several proximity dimensions should be taken into account when estimating the innovative power of a firm or industry. When the empirical findings are considered together it is evident that the local environment generates relational proximity between agents through formal- and informal networks. This proximity reproduces and rejuvenates the localized capabilities by allowing for the combination of heterogeneous pieces of knowledge in firms through local unrelated labor inflow. In conclusion, time and place are the paramount dimensions that shape the micro-dynamics of knowledge generation and innovation in firms.

O objetivo desta tese é a proposição de uma classificação para organizar a perspectiva multidimensional da proximidade. Essa classificação organiza formas para medição, representação e análise das múltiplas dimensões da proximidade. Isto dado que ainda não foram utilizados indicadores suficientemente satisfatórios para medir as várias dimensões da proximidade e que a forma como as medidas são definidas podem influenciar os resultados inferênciais obtidos nos modelos estatísticos. Este estudo levantou as definições conceituais para a proximidade e as suas principais formas empíricas de medição e propõe uma organização que possibilita a escolha entre as diferentes possibilidades de indicadores e de outras técnicas para interpretação e análise dessas medidas. Com isso foi realizada uma aplicação com dados da interação Universidade-Empresa no Brasil. A análise avaliou tanto os limites do ganho com a proximidade quanto os mecanismos de complementariedade e de substituição entre as dimensões geográfica e cognitiva. O foco na proximidade cognitiva favoreceu a comparação entre diferentes formas para a sua medição. Comparação realizada na perspectiva geométrica com uso de Escalonamento Multidimensional (EMD) e Análise de Correspondência (AC). Os resultados mostram a importância que precisa ser dada na forma como as medidas são definidas e representadas já que as interpretações dos resultados inferênciais nos modelos estatísticos estão diretamente relacionadas com esse passo inicial. As contribuições com esse estudo são de natureza metodológica, porém de caráter prático para guiar estudos futuros a partir das técnicas organizadas. Dentro desse escopo, a tese visa que outras técnicas, ainda pouco exploradas se apresentem como alternativas de uso para a ampliação da literatura da área. Por fim, a própria classificação pode ser adaptada para incluir outras técnicas analíticas que incorporem dados não espaciais ou que lidam com dados de não proximidade. / The aim of this thesis is the proposition of a classification to organize the multidimensional perspective of proximity. This classification organizes forms for measurement, representation and analysis of the multiple dimensions of proximity. This is because satisfactory indicators have not yet been used to measure the proximity and that the way in which the measures are defined can influence the inferential results obtained in the statistical models. This study raised the conceptual definitions for proximity and its main empirical forms of measurement and proposes an organization that allows the choice between the different possibilities of indicators and other techniques for interpretation and analysis of these measures. Hence, an application with data of the University-Industry interaction in Brazil was performed. The analysis evaluated both the limits of the gain with proximity and the mechanisms of complementarity and substitution between the geographic and cognitive dimensions. The focus on cognitive proximity favoured the comparison between different ways of measuring it. Comparison carried out in the geometric perspective using Multidimensional Scaling and Correspondence Analysis. The results show the importance that must be given in the way the measures are defined and represented since the interpretations of the inferential results in the statistical models are directly related to this initial step. The contributions with this study are of a methodological nature, but of a practical nature to guide future studies from the techniques organized. Within this scope, the thesis aims that other techniques, still little explored are presented as alternatives of use for the broaden of the literature of the area. Finally, the classification itself can be adapted to include other analytical techniques that incorporate non-spatial data or that deal with non-proximity data.

O objetivo desta tese é a proposição de uma classificação para organizar a perspectiva multidimensional da proximidade. Essa classificação organiza formas para medição, representação e análise das múltiplas dimensões da proximidade. Isto dado que ainda não foram utilizados indicadores suficientemente satisfatórios para medir as várias dimensões da proximidade e que a forma como as medidas são definidas podem influenciar os resultados inferênciais obtidos nos modelos estatísticos. Este estudo levantou as definições conceituais para a proximidade e as suas principais formas empíricas de medição e propõe uma organização que possibilita a escolha entre as diferentes possibilidades de indicadores e de outras técnicas para interpretação e análise dessas medidas. Com isso foi realizada uma aplicação com dados da interação Universidade-Empresa no Brasil. A análise avaliou tanto os limites do ganho com a proximidade quanto os mecanismos de complementariedade e de substituição entre as dimensões geográfica e cognitiva. O foco na proximidade cognitiva favoreceu a comparação entre diferentes formas para a sua medição. Comparação realizada na perspectiva geométrica com uso de Escalonamento Multidimensional (EMD) e Análise de Correspondência (AC). Os resultados mostram a importância que precisa ser dada na forma como as medidas são definidas e representadas já que as interpretações dos resultados inferênciais nos modelos estatísticos estão diretamente relacionadas com esse passo inicial. As contribuições com esse estudo são de natureza metodológica, porém de caráter prático para guiar estudos futuros a partir das técnicas organizadas. Dentro desse escopo, a tese visa que outras técnicas, ainda pouco exploradas se apresentem como alternativas de uso para a ampliação da literatura da área. Por fim, a própria classificação pode ser adaptada para incluir outras técnicas analíticas que incorporem dados não espaciais ou que lidam com dados de não proximidade. / The aim of this thesis is the proposition of a classification to organize the multidimensional perspective of proximity. This classification organizes forms for measurement, representation and analysis of the multiple dimensions of proximity. This is because satisfactory indicators have not yet been used to measure the proximity and that the way in which the measures are defined can influence the inferential results obtained in the statistical models. This study raised the conceptual definitions for proximity and its main empirical forms of measurement and proposes an organization that allows the choice between the different possibilities of indicators and other techniques for interpretation and analysis of these measures. Hence, an application with data of the University-Industry interaction in Brazil was performed. The analysis evaluated both the limits of the gain with proximity and the mechanisms of complementarity and substitution between the geographic and cognitive dimensions. The focus on cognitive proximity favoured the comparison between different ways of measuring it. Comparison carried out in the geometric perspective using Multidimensional Scaling and Correspondence Analysis. The results show the importance that must be given in the way the measures are defined and represented since the interpretations of the inferential results in the statistical models are directly related to this initial step. The contributions with this study are of a methodological nature, but of a practical nature to guide future studies from the techniques organized. Within this scope, the thesis aims that other techniques, still little explored are presented as alternatives of use for the broaden of the literature of the area. Finally, the classification itself can be adapted to include other analytical techniques that incorporate non-spatial data or that deal with non-proximity data.

Abstract The thesis is an investigation into the spatio-temporal knowledge creation of international research groups. Knowledge is spatial: it is created in geographical places, shared in communicative space and analyzed in the cognitive space of the mind. Knowledge is a process that evolves in the interactions and interpretations of people. It is created in the tempo of working days that is experienced as flowing or disrupted time. These dimensions of spaces and times come together as the contexts of knowledge creation. In current internationalizing universities, work groups increasingly consist of members from different countries and with varying professional backgrounds. In their endeavor to create globally renowned results, the international research groups are both challenged and enabled by their diversity. Multiple views can benefit the creation of novelty, but reaching a common understanding in a diverse group might not be easy. Current research calls for a more detailed understanding of the processes and contexts of knowledge creation in international groups. The empirical part of the thesis is comprised of two case study compilations. Case A focuses on the University of Oulu, with its foreign employees and three international research groups from the humanist, scientific and technical fields. Four international research groups from the fields of science and technology, led by top foreign professors and funded by the Finland Distinguished Professor Programme, form case B. Altogether, the main materials include 37 interviews, 123 weekly diaries, observation and a survey (67 replies). The main methods applied are mental mapping, discourse analysis and content analysis. According to the results, knowledge creation is a complex spatio-temporal process that often leads to unexpected results. Processes of interaction and interpretation enable the group to reach cognitive friction. This fertile ground of knowledge creation can be reached when a group possesses enough common understanding and enough diversity in their individual interpretations. Humanist, scientific and technical knowledges are created in multiple contexts through group-specific discourses. Knowledge itself is a spatio-temporal process of stages, flows and disruptions towards the not-yet-known. Knowledge is inseparable from individuals, groups, the processes and contexts of creation. / Tiivistelmä Väitöskirjassa tarkastellaan tiedon jalostamista tilassa ja ajassa kansainvälisissä tutkimusryhmissä. Tieto on tilallista: sitä jalostetaan maantieteellisissä paikoissa, jaetaan kommunikaation tilassa ja analysoidaan mielen kognitiivisessa tilassa. Tieto on prosessi, jota ihmiset jalostavat vuorovaikutuksen ja tulkinnan kautta. Tietoa jalostetaan aikarajojen ja työpäivien määrittelemässä ajassa, joka voi tuntua virtaavan tai pysähtelevän. Nämä tilan ja ajan ulottuvuudet yhdistyvät tiedon jalostamisen kontekstiksi. Kansainvälistyvissä yliopistoissa työryhmät koostuvat enenevissä määrin jäsenistä, jotka edustavat eri kansallisuuksia ja joiden ammatilliset taustat ovat erilaiset. Monimuotoisuus on sekä haaste että mahdollisuus ryhmien tavoitteelle luoda kansainvälisesti merkittäviä tuloksia. Monipuoliset näkökulmat voivat edesauttaa tiedon jalostamista, mutta yhteisymmärryksen muodostaminen ryhmässä voi olla haastavaa. Nykytutkimuksessa tarvitaan syvällisempää ymmärrystä tiedon jalostamisen prosesseista ja konteksteista kansainvälisissä ryhmissä. Väitöskirjan empiirinen osa sisältää kaksi tapaustutkimuskokonaisuutta. Tapaus A koskee Oulun yliopiston ulkomaalaisia työntekijöitä ja kolmea kansainvälistä tutkimusryhmää humanistiselta, luonnontieteelliseltä ja tekniseltä alalta. Tapauksen B muodostavat neljä kansainvälistä luonnontieteen ja tekniikan alan tutkimusryhmää, joiden johtajat ovat ulkomaalaisia huippuprofessoreita ja jotka ovat Finland Distinguished Professor -ohjelman rahoittamia. Yhteensä pääaineisto koostuu 37 haastattelusta, 123 viikkopäiväkirjasta, havainnoinnista ja kyselystä (67 vastausta). Aineiston käsittelyn päämenetelmät ovat mentaalikartta-, diskurssi- ja sisällönanalyysi. Tulosten mukaan tiedon jalostaminen on monimutkainen prosessi, joka usein johtaa odottamattomiin tuloksiin. Vuorovaikutuksen ja tulkitsemisen prosessit mahdollistavat kognitiivisen kitkan muodostumisen, joka on hedelmällinen maaperä tiedon jalostamiselle. Sen saavuttamiseksi ryhmässä tarvitaan riittävästi yhteisymmärrystä ja riittävästi monimuotoisuutta jäsenten yksilöllisissä tulkinnoissa. Humanistiset, luonnontieteelliset ja tekniset tiedot jalostetaan ryhmän sisäisissä diskursseissa monipuolisissa konteksteissa. Tieto on tilallinen ja ajallinen prosessi kohti toistaiseksi tietämätöntä. Tieto on erottamaton osa ihmisiä, ryhmiä ja jalostamisen prosesseja ja konteksteja.