2. MARCO TEÓRICO
2.3 COMPONENTES DE LA UNIDAD DE COILED TUBING
2.3.6 TUBERÍA FLEXIBLE
2.3.6.1 Característica de la tubería flexible
The data from the FGDs is analyzed using the method of content analysis. The content analysis enables structuring of the data as well as facilitates analysis and interpretation of reoccurring patterns (Harwood and Garry, 2003: 479). The content analysis is guided by the hypotheses and the analytical framework summarized in Section 2.5. A coding scheme is developed initially because it filters the information relevant for the research questions. The author defined the classification criteria, so-called a-priori categories (Mayring, 2010: 65), that are applied in a deductive manner to the text material. This approach towards content analysis is also termed as “Typisierende Strukturierung” (Mayring, 2010: 92-94; 98-101). Although the analysis is guided by a conceptual framework, other high frequency categories and relevant outliers referring to micro life insurance demand behavior are considered in the second step of the content analysis. This approach reflects the exploratory nature of the qualitative research paradigm. In the first step, the researcher started by developing a provisional coding frame. Displayed in Table 15 below, this coding frame consists of the a-priori categories reflecting the theory-induced demand factors.
Table 15: Content analysis –A-priori coding scheme
Category Sub-category Definition
Social capital Membership Coded if statement refers to a household’s membership in a formal social network7
Generalized trust Coded if statement refers to a person’s general attitude of trusting others
Creation of awareness / understanding
Coded if statement refers to an information exchange with others, a discussion about the topic of insurance or advice
Imitation Coded if statement refers to an imitative behavior in insurance consumption, i.e. insurance is bought because peer did Trust building Coded if statement indicates that
trust in insurance increases because of a peer effect, i.e. claims
experience of peers
Informal risk sharing Coded if respondent receives support from family network in case of an unexpected financial loss because of an uncertain event
Religion Interest Coded if statement refers to the involvement of interest in insurance Mutuality Coded if statement refers to the
solidarity character of insurance Risk aversion Coded if statement refers to a
person’s willingness to take risk
Economic factors
Income Coded if statement refers to a
person’s (labor) income
Wealth Coded if statement refers to a
person’s wealth, her assets or land ownership
Price of insurance Coded if statement refers to insurance premium or the price of
7 The definition of formal social networks follows the definition used in the quantitative study and includes membership in at least one of the following: village development society, women rural development society, religious society, savings group/association, agricultural association/fishery society, sports club, and political party.
insurance
Remittances Coded if statement refers to
remittances received or paid
Demographic factors
Bequest motive Children Coded if statement refers to a person’s number of children
Age Coded if statement refers to a
person’s age
Family status Coded if statement refers to a person’s family status
Gender Coded if statement refers to a
person’s gender
Social and cultural factors
Education Coded if statement refers to a
person’s education
Financial literacy Coded if statement refers to a person’s financial literacy or knowledge about insurance
Structural factors
Access to credit Coded if statement refers to a
person’s access and usage of credits
Access to savings Coded if statement refers to a person’s access and usage of savings
Social assistance Coded if statement refers to social assistance received
Contract non- performance
Refusal of claims payments
Coded if statement refers to an insurer who refused to settle a claim Delay in claims payment Coded if statement refers to an
insurer who delayed claims settlement
Solvency of the insurer Coded if statement refers to an insurer who might go bankrupt
Risk aversion and exposure
Risk exposure Coded if statement refers to a person’s current or previous risk exposure
Other risk management strategies
Reduced consumption Coded if statement refers to a person who manages her
unexpected losses by reducing her consumption
Borrowing Coded if statement refers to a person who manages her
unexpected losses by borrowing
Pawning Coded if statement refers to a
person who manages her unexpected losses by pawning
Savings Coded if statement refers to a person who manages her
unexpected losses by drawing from savings
Sale of assets Coded if statement refers to a person who manages her unexpected losses by selling her assets
Source: Author’s own
Furthermore, two attributes are added to every coded text passage for data analysis: (i) religious denomination: Hindu, Muslim, or Christian and (ii) the kind of insurance in focus: conventional insurance or Islamic insurance (Takaful), as given in Table 16. In the pilot coding of four FGDs transcripts (one policyholder and one non-policyholder group for the religion of Hinduism as well as for Islam), this provisional theory-led coding frame of the various categories and attributes was tested. To support a systematic coding process, the QSR International’s NVivo software was used.
Table 16: Content analysis –Attributes of provisional coding scheme
Attributes Characteristic Definition
Religious denomination Muslim Coded if respondent is Muslim or statement refers to a Muslim
Hindu Coded if respondent is Hindu or
statement refers to a Hindu Christian Coded if respondent is Christian or
statement refers to a Christian Type of insurance Conventional Coded if statement refers to
conventional micro life insurance Takaful (Islamic
insurance)
Coded if statement refers to Family Micro Takaful
Source: Author’s own
After the pilot coding, the coding frame was reassessed (Mayring, 2010: 50). So-called in-vivo codes or a-postiori categories were added, if the participants discussed demand factors beyond the a-priori defined categories (Barbour, 2007: 115-119). These categories, summarized in Table 17, were checked for their degree of differentiation in comparison with the a-priori categories and were clearly defined before addition to the coding frame.
The sub-category of ‘reputation’ was added to the category of social capital as Muslim participants expressed their concern that the purchase of conventional insurance could threaten their reputation within their community.
Four additional subcategories are developed for the category of religion: ‘importance of Islamic insurance provider and distributor’, ‘religiosity’, ‘living according to religious rules’, and ‘not specified’. The first sub-category includes text references to the importance of the insurance being offered by a service provider or distribution channel which runs its business in accordance with the Islamic principles. The sub-category ‘religiosity’ includes general statements about a person’s religiousness which cannot be attributed to one of the five dimensions of religiosity: intellect, ideology, public practice, private practice, and experience (Huber, 2007: 21), while the category ‘living according to religious rules’ refers to the specific dimension of religiosity of a person’s public or private practice. Several respondents related their insurance consumption decision to their personal behavior of living their lives in accordance with the rules of their religion. In some cases, text material related to religion was not specific enough to be clearly assigned to a sub-category. For example, some respondents stated that there is (not) a relationship between micro life insurance uptake and religion but did not provide any further information. Therefore, the sub-category ‘not specified’ was introduced. The participants frequently referred to the terms and conditions of the offered policies as a decision criterion for purchasing insurance. Hence, the category ‘terms and conditions’ was added to the coding frame. As the number of text references is considerably large, it was further classified into two sub-categories: (i) (dis)satisfaction with the ‘benefits’ of the insurance product and (ii) the ‘mandatory’ nature of an insurance product. The category ‘mandatory’ is applied, if the person either refers to vehicle and livestock insurance that is mandatory by law or credit life products where the customer is obliged to get insurance.
Several participants described their experiences with dishonest agents who defrauded them by pocketing their premium money and never sending it to the insurance company for issuance of an insurance policy in their names. To reflect this finding, an additional sub-category ‘fraud by agent’ was created within the main category of ‘contract non- performance’. As in the case of religion, further references were made by the
participants to the context of contract non-performance. However, the information was not detailed enough to assign the text material to a separate sub-category. Therefore, it led to the introduction of the sub-category ‘not specified’.
During the first coding cycle, it became obvious that errors in sampling occurred as outlined in the previous section. Usually, the groups consisted of both policyholders and non-policyholders, contrary to the initial decision to run separate focus groups. Hence, the author decided to change the data analysis strategy and introduced a third attribute to the coding frame. The purpose of this attribute is to allow for a distinction between the motivations and characteristics assigned to policyholders and non-policyholders. If a focus group participant stated a reason to support insurance consumption, the text was coded as a ‘driver’. Whereas, all the statements about a characteristic or motivation that hinders uptake of micro insurance were coded as a ‘barrier’. If the statement was neither positive nor negative towards insurance consumption, the author applied the code ‘not specified’. In a nut shell, the following categories and attributes were added:
Table 17: Content analysis –Additions to the a-priori coding frame
Category Sub-category Definition
Social capital Reputation Coded if statement refers to a person’s insurance consumption and its effect on her reputation within her social network
Religion Importance of Islamic insurance provider and distributor
Coded if statement emphasizes the importance of an insurance supply or distribution by an institution that operates based on Islamic principles Living according to religious
rules
Coded if respondent states that she follows religious rules in her life Religiosity Coded if statement refers to a
person’s religiosity in general without a specific reference to the
dimensions of intellect, ideology, public practice, private practice or experiences (Huber, 2007: 21) Not specified Coded if statement is related to a
person’s religion but not specific enough for further classification
Terms and conditions
Benefits Coded if statement refers to the
benefits of an insurance product Mandatory Coded if statement refers to a credit
life insurance product or a cover mandatory by law, such as vehicle or livestock
Contract non- performance
Fraud by agent Coded if statement refers to fraud by an insurance agent
Not specified Coded if statement is related to contract non-performance but not specific enough for further classification
Attributes Characteristic Definition
Insurance uptake Driver Coded if statement is related to a
factor supporting micro life insurance consumption
Barrier Coded if statement is related to a
factor preventing or reducing micro life insurance consumption
Not specified Coded if statement is related to micro life insurance consumption but its effect is unclear
Based on the revised coding framework, a coding cycle followed which included all the transcribed text materials. After the first round of coding all the focus group transcripts, the author of this study reviewed all the categories and the text materials assigned to them. If no text material could be assigned to a theory-led category, the category was dropped. This was the case for five sub-categories related to the research questions on social capital and religion including: (i) ‘membership’, (ii) ‘generalized trust’, (iii) ‘trust building’, (iv) ‘risk aversion’, and (v) ‘mutuality' and another six categories suggested by the analytical framework: (i) ‘remittances’, (ii) ‘age’, (iii) ‘gender’, (iv) ‘access to credit’, (v) ‘savings’, and (vi) ‘social assistance’. The categories with fewer references were assessed for mergers. Here, the statements assigned to the sub-categories ‘refusal of claims payments’ and ‘delay in claims payments’ were merged and the sub-category renamed as ‘payment of claims’. Further, sub-categories for ‘other risk management strategies’ are combined. In other cases, the definitions of codes were fine-tuned and the categories were adapted, wherever necessary. For the case of the category ‘number of children’, most of the coded text passages referred to the motivation of purchasing insurance for the protection of children and not to the number of children. Hence, the category was renamed as ‘protection for children’. The category ‘price of insurance’ was labeled ‘mode of premium payment’ and moved to the category ‘terms and conditions’. To ensure an unambiguous data analysis and due to the similar nature of the two categories, ‘wealth’ and ‘income’ were merged as one category.
Moreover, the categories and their assigned text passages were reviewed for consistency. If coding errors were found, the material was recoded. All the uncoded materials were cross-checked and appropriate categories and attributes were assigned to the text passages, wherever possible.
A second coding cycle was exercised based on the revised coding framework. Lastly, the accuracy of the coded text material was double-checked during the analysis. If results appeared inconsistent or not plausible, the coded materials were reviewed. If a coding error had occurred, the transcript excerpt was recoded.
Table 18 shows the frequency counts of the final coding scheme. It also presents the different categories, sub-categories and if applicable, the number of statements made during the various FGDs and the number of sources. In some cases, subtotals presented
for a category are not equal to the sum across the sub-categories. This is caused by the quotes assigned to two subcategories, but only accounted once for the category to avoid double-counting. The next section evaluates the quality of the content analysis.
Table 18: Content analysis –Final coding scheme and frequency of statements
No. of references No. of
sources
Categories Sum Driver Barrier Not specified
Social capital 25 9 5 11 6
Awareness/ understanding 18 7 0 11 5
Informal risk sharing 4 1 3 0 3
Reputation 2 0 2 0 1
Imitation 1 1 0 0 1
Religion 77 27 33 17 8
Religiosity 6 4 2 0 4
Living according to religious rules 13 2 11 0 4
Interest 31 9 17 5 5
Importance of Islamic insurance provider or distributor
15 9 2 4 4
Not specified 13 3 2 8 6
Income & wealth 56 15 40 1 9
Bequest motive 17 14 3 0 6
Children 14 13 1 0 6
Family status 3 1 2 0 3
Education 7 0 2 5 3
Financial literacy 7 4 3 0 4
Other risk management strategies 53 0 3 50 9
Contract non-performance 16 0 16 0 5
Payment of claims 8 0 8 0 5
Fraud by agent 5 0 5 0 4
Not specified 3 0 3 0 5
Risk exposure 7 4 3 0 5
Terms and conditions 38 30 6 2 7
Benefits 31 24 5 2 7
Mandatory 6 5 1 0 4
Mode of premium payment 3 1 2 0 3
Source: Author’s own
Note: In some cases, subtotals presented for a category are not equal to the sum across the subcategories. This is caused by the quotes assigned to two subcategories, but only accounted once for the category to avoid double-counting.