You are on page 1of 10

Research strategy and design: Case study 1

The effectiveness of computer-based training at Falcon Insurance Company

(The techniques used in this case may appear complicated when you first read about them. Do not worry; the case is included as an example of a research strategy, not techniques. Further details on the techniques of observation, in-depth interviews and questionnaires can be found in Chapters 8, 9 and 10 respectively.) Ingrid works for Falcon Insurance, a large UK insurance company offering insurance in the life and motor markets. As part of its training portfolio Falcon Insurance was considering the introduction of computer based training (CBT) for junior-level courses such as telephone manner. In order to make an informed decision about introducing CBT, Falcon needed to answer the question: Is CBT an effective training medium? After discussion with her project tutor Ingrid decided this would make a good research project. An extensive literature review enabled her to define the term effective and generate four interrelated research questions: y y y y Is CBT effective for individual learning? Is CBT effective for group study? Is learning in groups more effective than self-study? Does a match of training media to learning style improve learning?

Ingrid wanted the research to provide both quantitative results, which described the outcomes of CBT in relation to other forms of training, and qualitative results, which explored trainees opinions and behaviours and the meanings behind these outcomes. In order to describe and explore the effect of CBT on learning Ingrid decided to undertake a study that examined the impact of this type of training over an extended period. The independent variable was the training media used, and the dependent variable was the learning that took place. A review of the literature suggested that four distinct groups would be needed: 1. self-study CBT: using the CBT software individually without a trainer to facilitate and provide input; 2. group study CBT: using the same CBT medium but studying as a group of trainees without a trainer to facilitate and provide input; 3. traditional training course: using the message and content of the CBT but applying it to a facilitated group in which no technology was used and the input was provided by a trainer; 4. control group: where no training took place during the study period to provide a bench mark. A representative sample of 120 of Falcon Insurance s employees was selected and 30 were allocated to each group at random. Data were collected using a combination of questionnaires, observation and unstructured interviews (Table 1).

Table 1 The research strategy Group Action taken: Immediately before training 1 Self-study CBT Questionnaire and learning styles Questionnaire During training Descriptive observation Immediately after training Questionnaire and separate in-depth interview Six months after training Questionnaire

2 Group study CBT

Questionnaire and learning styles Questionnaire Questionnaire and learning styles Questionnaire Questionnaire

Descriptive observation

Questionnaire and separate in-depth interview Questionnaire and separate in-depth interview


3 Traditional training

Descriptive observation


4 Control group


Ingrid decided to use pre-tests and post-tests on each of the four distinct groups. These data would be collected using a questionnaire on three separate occasions: immediately before training, immediately after training, and 6 months after training. A confidential self-administered questionnaire, to measure quantitatively any changes in individuals knowledge and understanding over the study period, was designed and pilot tested. Questions tested the respondents knowledge and understanding that should have been developed by the training they would receive. The control group was also tested, despite receiving no training, to measure any change in knowledge and understanding due to other factors. In addition, questions were included to measure the trainees reactions to various aspects of the training as well as attribute questions to assess the representativeness of the sample. The questionnaire was administered to all four groups for the pre-test and, with minor alterations, the post-test stages. In addition, all respondents who actually undertook some form of training were asked to complete a learning styles questionnaire. An important part of Ingrid s research was to examine the social context of the groups studied and to provide additional data with which to compare and explore the results of the study. Descriptive observation was used to collect data on the behaviour of each of the three groups of trainees, paying particular attention to events and observed actions and behaviours. Observers were given a sheet on which they recorded this information in a structured manner using a predetermined scheme. This scheme was developed to minimise the impact of subjectivity and inconsistency on the recorded response. Observations were

recorded while observing each of the three trainee groups, the observer remaining only as an observer and not engaging with the group. Ingrid used unstructured interviews to collect views about group learning from a stratified sample of trainees immediately after the training. In particular, she asked trainees to discuss their views on learning in a group and whether they preferred this to self-study. Trainees who had received group study CBT were asked to discuss their views on the effectiveness of the medium in the group context and about the effect of group interaction in their particular session. Trainees were asked to consider whether or not they would put their learning into practice in their jobs. Finally, the trainees were asked to consider the match between their preferred learning style and the training they had received, and whether it had improved their learning. Data collected from the questionnaire, observation, and unstructured interviews were subsequently analysed.

Questions 1 Which type(s) of research strategy are employed in this study? 2 What are the benefits of adding a longitudinal dimension to this research? 3 What are the benefits of using multiple methods of data collection?

Saunders, Lewis amd Thornhill: Research Methods for Business Students, 5th edition, Additional Case Studies

Research strategy and design: Case study 2 (Group discussion)

Article 1: International Financial Reporting Standards and the quality of financial statement information by George Iatridis, International Review of Financial Analysis 19 (2010) 193 204. Introduction (extract) Agency theory makes a number of predictions regarding the behaviour of managers. It would suggest that by adopting IFRSs, firms act optimally and promote financial reporting quality and investor interests (see Fields, Lys & Vincent, 2001). For example, highly leveraged firms would be keen to adopt IFRSs in order to satisfy the needs of lenders and the requirements of debt covenants and/or avoid political attention and scrutiny (see Lambert, 2001). Healy (1985) suggests that the flexibility allowable in financial reporting may cause managers to behave opportunistically (see Burgstahler & Dichev, 1997; Weil, Fung, Graham & Fagotto, 2006). This would imply that managers might manage the reported earnings in order to demonstrate a favourable transition to IFRSs (DeFond & Park, 1997; Hand & Skantz, 1998; Fields et al., 2001) or to avoid adverse implications on their profit figures and debt-paying ability (Watson, Shrives & Marston, 2002). A crucial question is how the adoption of IFRSs would affect accounting quality. The fact that the UK is considered to be a common-law country with active stock and debt markets, a diverse base of investors, strong investor protection mechanisms and investor-oriented financial reporting (Tendeloo & Vanstraelen, 2005) might have benefited the transition process. Previous studies (Tarca, 2004; Barth et al., 2005; Lang et al., 2005; Tendeloo & Vanstraelen, 2005; Hung & Subramanyam, 2007) exhibit that IFRSs are information-oriented and improve the quality of financial reporting, thereby meeting the information needs of investors and reinforcing the structures of the stock market. The study focuses on firms listed on the London Stock Exchange and determines whether the adoption of IFRSs has improved accounting quality. The implementation of IFRSs is compulsory for listed firms that belong to member states of the European Union (1606/2002/EC), the effective date being 1 January 2005. The study is exploratory in nature and tests for systematic differences among UK firms in the light of IFRS adoption. The period under analysis is the official period of adoption, i.e. 2005, and the pre-official adoption period, i.e. 2004. The empirical investigation concentrates in the examination of the earnings management potential under IFRSs and the value relevance of IFRS-based accounting reports as compared to UK GAAP. The study shows that IFRS adoption reduces the scope for earnings management, leads to more timely loss recognition and to more value relevant accounting measures. The motivation of the study relates to whether a particular accounting method/rule has a real impact, such that managerial decisions would need to be altered to minimise the adverse effects of the accounting change. The adverse impact of an accounting change could be mitigated by smoothing the accounting numbers and manipulating the accounting change to suit the financial decisions of the firm (see Bazaz & Senteney, 2001). Such information would be useful for the accounting standard setting process, particularly with regard to whether stricter or more flexible financial reporting should be imposed (see Levitt, 1998). There have been a number of studies about IFRS implementation in different national settings. For example, Eccher and Healy (2003) study IFRSs and the Chinese GAAP, while Leuz (2003) and Hung and Subramanyam (2004) study IFRSs and the German GAAP. The UK is a common-law country with strong investor protection mechanisms in place, while for example Germany and China are code-law countries. Hence, it would be fruitful to investigate the situation for such a setting and to identify the impact of IFRS adoption on accounting quality, taking into consideration that the UK GAAP is a common-law and investor-oriented system.

Article 2: Multinational banks from developing versus developed countries: Competing in the same arena? by Andreas Petrou, Journal of International Management 13 (2007) 376 397. Introduction A phenomenon of the globalization of the world economy is the increasing foreign direct investment of companies based in developing countries. The emergence of global competitors of the likes of Haier, Lenovo and Cemex raises questions about how firms from developing countries compete with rivals from developed countries, a research issue identified by a number of international business scholars (Li, 2003; Ramamurti, 2004; Yeung, 1994). Most researchers studied Multinational Enterprises (MNEs) without considering whether they originate from a developed or a developing country assuming that these two groups of MNEs are similar (Caves, 1982; Chen et al., 2004; Dunning, 1980). However, studies of MNEs from developing countries suggested that these multinationals have certain differences compared to MNEs from developed countries; they are smaller, less international and may not possess ownership advantages during their first steps of internationalization (Giddy and Young, 1982; Mathews, 2002; Van-Hoesel, 1999). Given that FDI theories suggest that possession of ownership advantage is a requirement for internationalization (Hymer, 1976; Kindleberger, 1969), this raises the question of why MNEs from developing countries internationalize in the first place. This may also suggest that, more research on this issue is required before we start creating consensus (Makino et al., 2002). In addition, studies of MNEs from developing countries concentrated primarily on manufacturing firms as researchers studied mostly MNEs from South East Asia and China (Makino et al., 2002; Lecraw, 1993; Li, 2003; Van-Hoesel, 1999). Service firms which have different characteristics from manufacturing firms such as simultaneous production and consumption and perishable products which often constraint internationalization options have not been studied from this perspective despite the fact that internationalization motives may be different. For example, a major internationalization motive of service firms is to follow their clients which have not been shown to be a major driver of manufacturing firm internationalization (Akbar and McBridge, 2004; Enderwick, 1989; Erramilli, 1991). The purpose of this study is to compare directly multinational service firms from developing countries with similar companies from developed countries in order to understand how MNEs from developing countries compete in a globalized environment with fewer resources and less international experience. We are particularly interested to understand a) differences in foreign market strategic motives; b) differences in location choices and location entry strategies. Ownership, location and internalization related motives identified in the literature and patterns of foreign investment will be compared across the two groups in order to establish differences that may explain different FDI behaviour. Furthermore, managers' strategic motives will be measured directly rather than using proxy variables, unlike most FDI studies, because it is difficult to capture perceptual constructs like motives, with proxy variables (Cyert and March, 1963; Makino et al., 2002). In order to control for service industry differences (Boddewyn et al., 1986) this study will focus on commercial banking. This industry is more regulated which results in differences in local markets, and assets are mostly financial which increases the cost of failure. Furthermore, this is an important topic for banks as the last two decades we saw a significant increase in internationalization from banks from developing countries. For example, HSBC started as a small bank in East Asia to become a major force in global banking by pursuing more than twenty foreign market entries the last two decades in countries like UK, US, France, Turkey, Brazil and India, which raises the question of how could a bank from a developing country with little international exposure such as Bank of China become a credible international competitor. In addition, cross-border activity in Europe such as in

Eastern Europe, in the Balkans, in the Baltic countries as well as in main markets such as Italy and Germany provides the grounds for increased international competition among banks from developed and developing countries.

Group discussion to answer all or some of these questions: 1. What strategy do you suggest for this study (e.g. quantitative, qualitative or mixed) and why? 2. What design do you suggest for this study (e.g. survey, case study, cross-sectional, archival, etc.) and why? 3. What possible sample do you suggest and why? 4. What kinds of data do you suggest to collect, i.e. independent variable(s), dependent variable(s), control variable(s)? 5. What potential databases for data collection, if you know? 6. Other suggestions?

Research strategy and design: Case study 2 (Real research designs in the articles discussed)

Article 1: International Financial Reporting Standards and the quality of financial statement information by George Iatridis, International Review of Financial Analysis 19 (2010) 193 204. Research structure Datasets and empirical methods The analysis focuses on firms that adopted IFRSs. The effective date for the adoption of IFRSs for listed firms that belong to member states of the European Union is 1 January 2005. The empirical analysis concentrates on the official adoption period of IFRSs, i.e. 2005, and the pre-official adoption period, i.e. 2004. The sample consists of 241 UK firms. All sample firms adopted IFRSs in the official adoption period. Prior to IFRSs, all sample firms had been using the UK GAAP. Accounting and financial data were collected from DataStream. Information about the accounting policies of the sample firms was obtained from their financial statements, which were collected from the Financial Times Annual Report Service. All sample firms are listed on the London Stock Exchange. The analysis has excluded banks, insurance, pension and brokerage firms, as their accounting measures are not always comparable with those of industrial firms. Appendix A presents the industrial sector structure of the sample firms. Appendix B shows the explanatory variables that are employed in the empirical analysis. Appendix C presents the institutional differences between IFRSs and the UK GAAP. The research hypotheses are tested using the binary logistic regression analysis and the OLS regression analysis. The logistic regression is useful in analysing categorical data, where the dependent variable is dichotomous and takes only two values, i.e. 0 and 1. The parameters of the logistic regression are estimated based on the maximum likelihood method, while the hypothesis testing is based on the Wald statistic. The diagnostic tests entailed an assessment of: (i) the relative significance of the estimated coefficients (p-value 0.01; two-tailed); (ii) the magnitudes of the logit models' Studentized residuals (b3.0); and (iii) the naive proportional chance model (see Joy & Tollefson, 1975). All the logistic regression results reported in this study have consistently passed those tests. The study has accounted for heteroscedasticity, autocorrelation, departure from normality and multicollinearity, where appropriate. The tests that have been performed to check the OLS assumptions are the White test and the Autoregressive Conditional Heteroscedasticity (ARCH) test for heteroscedasticity; the Durbin Watson test and the Breusch Godfrey test for autocorrelation; the Jarque Bera test for the departure from normality of residuals; and the correlation coefficients among the test variables for multicollinearity. ----------------------------------------------------------------------------------------------------------------------------------Notes (tutor added): Appendix A contains 21 industries such as Chemicals, Construction and building materials, Electricity, Engineering and machinery, etc. Appendix B shows the explanatory (independent) variables including Annual stock return, Book value per share, Cash flow per share, Discretionary accruals, Dividend pay-out, Dividend per share,, Dividend per share growth, Earnings per share, Interest cover, Large negative losses, Natural logarithm of market value, Long-term liabilities to capital employed, Net profit margin, Net profit to share price, Net profit per share, Operating cash flows scaled by total assets, Operating profit margin, Share price, Price to earnings growth, Return on asset, Small positive profits, Net profit scaled by total assets, Total liabilities to shareholders' funds, Change in operating cash flows scaled by total assets, and Change in net profit scaled by total assets.

Article 2: Multinational banks from developing versus developed countries: Competing in the same arena? by Andreas Petrou, Journal of International Management 13 (2007) 376 397. Methodology (extract) Sample Data were collected through a mail survey sent to 562 executives from 385 multinational banks identified in the Banker's Almanac. Questionnaires were sent to one or more executives in a bank considering the number of foreign operations of the bank; for every five foreign operations one respondent was selected (maximum five respondents). This approach has been employed by a number of empirical studies of internationalization (Agarwal and Ramaswami, 1992; Erramilli, 1991). Respondents were asked to provide information about a recent foreign entry. A total of 112 usable questionnaires were collected representing 93 banks which compares favourably with other perceptual studies of MNE internationalization. One of the main issues relating to post hoc research designs is that respondents may have difficulty accurately responding to events from their past unless they were either outstanding or recent (Emory, 1980). Besides encouraging executives to provide information about a recent foreign venture, respondents were carefully selected to ensure that they participated in foreign entry decisions by identifying senior managers from corporate strategic planning, the international division and members of the executive committee. Furthermore, data were tested for consistency by comparing information about nine ventures, which were provided by two different respondents, across all 7-point scale responses. It was found that 89% of responses were within an interval or less which is within the guidelines provided by Shortell and Zajac (1990). In order to evaluate non-response bias thirty non-respondents randomly selected were encouraged to complete a shorter version of the questionnaire. Responses and organizational demographics of sixteen participants were compared with the rest of the sample, which indicated no statistically significant differences on any of these dimensions. In addition, respondents to the first mailing were compared to respondents to the second mailing across the same organizational demographics. Again no statistically significant differences were found which might indicate that non-response bias is not present in the sample (Armstrong and Overton, 1977). Dependent variable In order to test differences between banks from developing countries and banks from developed countries two groups are formed where MNBs from developing countries are classified as 1 and those from developed countries as 0. In order to classify a country as developed or developing two criteria were employed considering a country's level of development and the degree of sophistication of a country's banking system respectively. The first criterion is approximated by a country's GDP with the threshold set at $25,000. It is recognised by IMF and the World Bank that this level of GDP is sustainable only if companies in the country are able to produce superior products and services which are competitive on a global basis. This may suggest that banks in the country are also sophisticated as they serve these demanding clients. Countries rich in resources such as Arab Emirates and Brunei were excluded from the developed country list. The second classification criterion was approximated by considering whether a country was represented by a top-100 bank as rated by the Banker's Almanac. These banks tend to be more resourceful and sophisticated which reflects the level of development of the banking system in the home country. The countries represented by the top-100 banks almost coincide with the GDP criterion with the exception of China which is a low GDP country but it is represented by top-100 banks, and some small and high GDP countries such as Ireland, Luxemburg and Norway which are not represented by a top-100 bank. Given that only China and Ireland are represented in our sample, the two criteria yield similar results. We choose to use the GDP criterion given that Chinese banks are state-own monopolies which it doesn't necessary imply banking system sophistication. Countries that are classified as developed 8

include some European Union countries, the US, Canada, Japan and Australia. In the sample there are 36 foreign market entries from banks based in developing countries and 76 entries from banks from developed countries representing a total of 38 countries as shown in Table 2. Independent variables Independent variables capture managers' perceptions at the time of the investment and therefore, psychometric measurement based on multiple items seemed to be the most appropriate approach for operationalizing these variables (Churchill, 1979; Peter, 1979). Most empirical studies use singleitem proxy exploratory variables, which often are rough approximations of the factors under investigation. This type of operationalization cannot fully capture the managerial aspect of strategy at the time of decision-making. This view is supported by the organizational behaviour literature where there is wide support for the importance of managerial perceptions in decision-making (Cyert and March, 1963). As no established scales with proven psychometric properties exist to measure the constructs of interest in banking, it was necessary to develop indicators that could represent the domain of each construct following Churchill's (1979) construct development paradigm. Initially a set of items for each construct was developed drawing on the existing literature, which were later discussed in two-hour sessions with nine senior bank executives who were familiar with entry mode selection. After data collection an iterative procedure between factor and reliability analysis was adopted in order to refine the set of indicators for each construct (Hair et al., 1998). Finally, factor scores were calculated for each construct by averaging the responses across indicators. The constructs emerged from this exercise are reliable with Cronbach's alpha exceeding 0.70. Follow the client is a two-item reliable scale with Cronbach's alpha equal to 0.805, which captures the bank's motivation to serve existing clients and markets from the home country. These two variables that were identified by studies of bank FDI (Fisher and Molyneux, 1996; Sabi, 1988) were also suggested by bank executives. Strategic selection of locations was developed based on a measurement developed by Kim and Hwang (1992). Respondents were asked to rate on a scale of 1 7 the following motives; establishing a strategic outpost for future expansion, sourcing of funds and entering a strategic market. These motives were found to influence bank location choices (Hoschka, 1993; Yannopoulos, 1983). For example, Yannopoulos (1983) argued that multinational banks require a sizeable and stable fund base which they can lend to business clients. As a result, MNBs source these funds by establishing operations in key currency markets. The construct developed was reliable with Cronbach's alpha equal to 0.781. Entry mode was operationalized by asking respondents to identify the percentage of shares owned of the foreign venture under discussion at the time of the investment. Following from previous studies wholly-owned is defined as an investment where the parent firm owns more than 95% of the equity (Chen andHennart, 2002; Gomes-Casseres, 1989), and partially-owned is defined as any entry where equity is shared with one or more partners and the bank owns between 5% and 95% of the equity (Chen and Hennart, 2002; Delios and Beamish, 2004; Stopford andWells, 1972). Control variables Ownership and internalization advantage related motives were used as control variables in order to represent all FDI theory dimensions. Empirical studies capture firm Ownership advantage by measuring advertising intensity and/or R&D spending. In banking like most service industries competition takes place around reputation, customer knowledge and ability to market products to target markets (Clarke et al., 2003; Tschoegl, 1987; Yannopoulos, 1983) and therefore, R&D spending contribute less to a bank's ownership advantage. Erramilli et al. (2002) studying service firms developed a construct to measure competitive advantage in attracting customers which is 9

adopted in this study. Respondents were asked to evaluate on a scale of 1 7 the importance of three motives; capitalizing on the bank's knowledge of the target market, taking advantage of the bank's reputation in the target market and capitalizing on the bank's marketing skills. The construct developed was reliable with Cronbach's alpha equal to 0.748. Synergies were operationalized through a five-item scale based on a construct developed by Kim and Hwang (1992). However, instead of identifying synergies in functional areas like manufacturing and marketing which bank executives considered inappropriate, a list of areas of bank synergies identified by Gray and Gray (1981) were adopted by asking respondents to rate on a scale of 1 7 expected benefits from the sharing and flow of resources among the bank's existing operations and the new venture in each of the following areas; investments, global asset liability management, commercial intelligence, account management and intra bank fund transfers. The construct developed was reliable with Cronbach's alpha equal to 0.812. Bank capabilities are also used as control variables; bank size and international experience were measured directly by the bank's tier one capital (Sabi, 1988) and the number of operations in foreign markets (Hennart and Park, 1993) respectively.