Social power of preadolescent children on influence in their mothers’ purchasing behavior: Initial study in Peruvian toy stores

Abstract

Purpose

This paper aims to investigate the relationship between legitimate and expert social power types of preadolescent children on the influence perception in their mothers’ purchasing behavior in Peruvian toy stores. The literature review takes into consideration the concepts of social power and the influence on family behavior to then focus on social power within family behavior with the purpose of mainly developing four hypotheses regarding purchasing behavior.

Design/methodology/approach

The methodology followed a non-experimental transversal correlational-causal design. A pilot sample size of 67 cases was used. The sample was based on an objective population of Peruvian mothers of families that live in northern Lima and that go to purchase toys to major shopping centers with their children aged 8-11 years.

Findings

The results show that the expert social power, as well as the legitimate social power, has a strong relationship. In addition, both social powers have an impact on the influence perception in purchasing child-mother, but not on the influence perception in purchasing mother-child. Moreover, the test of moderation of the expenditure level on toy purchases did not have an effect on the context that was studied.

Originality/value

The contribution shows that important changes are happening in the consumption behavior on the aspect of children influencing mothers, and that for Latin American contexts, the level of expenditure still does not crucially affect the causality demonstrated.

Citation

Carrillo, M., Gonzalez-Sparks, A. and Salcedo, N.U. (2018), "Social power of preadolescent children on influence in their mothers’ purchasing behavior: Initial study in Peruvian toy stores", Journal of Economics, Finance and Administrative Science, Vol. 23 No. 45, pp. 150-166. https://doi.org/10.1108/JEFAS-01-2018-0018

Introduction

Children should be seen as three markets in one (Aldea and Brandabur, 2012; McNeal, 1999); the actual market that spends money on their desires, the potential market for the majority of goods and services and an influence market that motivates their parents to consume different stuff. Children not only learn by copying their parents’ consumption behavior (Turner et al., 2006) but also apply pressure in the opposite direction; they influence their parents’ final purchasing decision behavior in three basic categories: toys, clothes and food (Nicholls and Cullen, 2004). Specifically, mothers are more likely to impulsively buy toys, clothing and sweets for children (Turčínková et al., 2012).

Likewise, the socialization of the consumer as “processes through which young people acquire abilities, knowledge, and relevant attitudes for their functioning as consumers in the marketplace” (Ward, 1972) leads to many children developing their own opinion and likes over the products that they want to buy (Turner et al., 2006).

Alonso and Grande (2013) argue that children influence purchasing in two manners. They induce the consumption of goods specific for them. In addition, we condition the purchases of goods in which children participate as consumers together with adults. They are able to influence the purchases because they remember the existence of the products and act as prescribers. Furthermore, it is said that children influence the purchase because they retain the product advertising messages better than their parents. Thus, in this way, they can motivate the product purchasing. “It has been confirmed that the qualitative composition of the basket of goods is different when it is done in the presence of children at the point of sales” (Alonso and Grande, 2013).

Research has found that children have power over their parents in family consumption decisions. McNeal (1999) estimates that children between ages of 4 and 12 influence on approximately US$188,000 annually on purchases related to the family in the USA. While in Peru, as per a study undertaken by the consulting company CCR for El Comercio (2014), children influence on the purchase of 62 per cent of Lima households. The toy market moves approximately between US$75 and 80m (Andina, 2010). Consequently, the influence of children on the family consumption decisions is a topic that merits attention.

Being so, what is the relationship between the legitimate social power (LSP) and expert social power (ESP) types of children on the influence perception of their mothers in Peruvian toy stores? This paper concentrates on Peruvian mothers living in the city of Lima, who have and live with their children who are in the preadolescent stage of development, meaning between 8 and 11 years. Mothers were selected for this research for two main reasons. First, mothers are more often the receptors of influence attempts than fathers (Cowan and Avants, 1988; Cowan et al., 1984). Second, mothers are usually the agents for family purchases and are considered to be better familiarized with the purchase influence attempt of their children.

For methodological purposes, the last studied relationship regarding passive social power of children on influence from the mothers’ perspective (Flurry and Burns, 2005) was used. To assess these variables, we applied instruments already validated by their authors and which have been used in other studies.

For these reasons, discussions in this field suggest extending this investigation (Flurry and Burns, 2005; Goodrich and Mangleburg, 2010): first, validating the veracity of the proposed constructs (Flurry and Burns, 2005); second, taking other populations as emerging markets; and third, considering the effect of social power at stages of influence in purchasing (Goodrich and Mangleburg, 2010).

Social power

Weber (1962) defines social power as “the probability that an actor within a social relationship is in the position of doing as he wills not withstanding resistance, independently over the basis on which this probability stands”. Likewise, Cartwright (1959), after taking Lewinian’s theory (Lewin, 1951), established the definition of power as “the induction of forces by entity B over A and another to the resistance of this induction created by A”. Therefore, in accordance with this conceptualization, it is the degree to which agent B has control over A’s behavior depending on the magnitude of force that B can exert over A and over A’s resistance.

The theorists of social power of the 1950’s and 1960’s have differed over the diverse aspects of the conceptualization of social power. However, Smith (1970) says that these theorists agree in two basic concepts of the definition. First, social power is the potential of one person to exert a force toward change in another person. Second, social power is not simply based on one quality or qualities that the powerful person possesses rather by complex conditions that rule the interdependence of the persons in a social relationship. Additionally, Elias (2008) mentions that since the study by French and Raven (1959), power is quantified in the capacity of a power possessor to persuade an identified objective, being the maximum influence possible that he or she can exert.

French’s (1956) taxonomy is composed of five types of power: reward, coercive, legitimate, expert and referent. The first one, reward, is “based on the perception of A, where B has the ability to mediate rewards for A”. The coercive power is “based on the perception of A, where B has the ability to mediate punishments for A”. The legitimate power is “based on A’s perception where B has a legitimate right of prescribing the behavior for A”. The expert power is “based on A’s perception, where B has a special knowledge or expertise”. Finally, the referent power is “based on the identification of A with B”. Then, this included a sixth power: informational or persuasion (Raven, 1993).

Social power in family behavior

The previously explained theory also considers the fathers and the children in a relationship of interdependence with a different level of power. When there is a conflict between the children and the parents’ perspective with respect to a consumption decision, for example, the possibility of purchasing a product, what brand to buy, how much to buy, among others, the children could strategically use their power to persuade the parents, thus gaining influence in the decision-making process (Cowan and Avants, 1988; Cowan et al., 1984; Kim et al., 1991).

As per Flurry and Burns (2005), the theory of social power suggests that the five bases can be used in family behavior in two ways: active and passive. Using power to influence is commonly considered active, although sometimes it can be passive, like when the mere presence of the power is influential (Podsakoff and Schriesheim, 1985; French and Raven, 1959). Both forms of power, active and passive, contribute to the potential of a person to achieve a result in accordance to his or her own reference. Hence, Flurry and Burns (2005) and Williams and Burns (2000) argue that children exercise influence through active social power and/or passive social power.

When there is no evidence of spoken words or manifested actions by the child, it is said that the exerted influence by him or her can also be passive. Consequently, the passive sources of power only need to be possessed to have an effect (Corfman and Lehman, 1987). Then, for a child, a source of power is passive if his or her parent infers its presence and acts upon it instead of upon any action by the child. This is known as child’s influence on the parents or the parent perceptions of the child’s undeclared preferences (Wells, 1965). As these children grow, they influence family purchasing decisions in a more passive manner, as the parents learn the likes and dislikes of their children and make buying decisions based on that (Roedder, 1999).

Being so, when talking about expert power, this means providing knowledge and superior ability to the influenced person, for which the child can possess more knowledge in determined categories of products such as clothing and toys, among others (Flurry and Burns, 2005). On the other hand, when talking about legitimate power, one perceives the right to control the opinion or behavior of the other person (Flurry and Burns, 2005). This power is derived from a justifiable right (Elias, 2008). A child has legitimate power when he perceives that he has the right to make a decision based on his or her interests. Being so, the following hypothesis arises:

H1.

There exists a direct positive relationship between passive expert social power and passive legitimate social power of preadolescent children as mothers’ perception in the toy stores context.

Influence in family purchasing behavior

While social power is “the potential influence of one person over another one” (Cartwright and Zander, 1968; Cartwright, 1965), Swasy (1979) defines the influence as a change in cognition, attitude, behavior or the emotion of a person.

Influence is defined as the use of power to achieve a result (Coleman, 1973). At the same time, influence could only be attained as a result of a reciprocal exchange process between two or more parties (Sprey, 1975). Olson et al. (1975) claim that the use of power exerts an influence known as “circular causal process”, for which the resulted power in any of the parts is a fusion of all perspectives of the parts involved in the decision-making process.

Flurry and Burns (2005), quoting Olson et al. (1975), say that given the reciprocal nature of the influence, if measuring the influence is wanted, it is necessary to measure the perspectives of all the significant members of the decision-making process. Also, French and Raven (1959) say that it is to be expected that the perceptions of the parts will be similar but not identical. That is why, it is important to consider that in the families that face a social power that could be exerted by a child, the mother could very well influence on its decision. For this reason, the following is considered as the second hypothesis:

H2.

The passive social power, expert and legitimate of preadolescent children have a direct effect on mother’s perception of the child’s influence (from mother to child) in a toy purchase.

However, demographical and structural changes in the households have changed children’s roles in families’ buying activities, increasing both their participation in family decision-making processes and their purchasing power (Flurry and Burns, 2005; Williams and Burns, 2000). In the present time, in many societies, both parents work. Therefore, Sellers (1989) states that the parents, with time limitations to go shopping, permit or even encourage their children to participate in the process of making decisions.

Wimalasiri (2000) says that influence is a term used to describe the interaction between the parents and their children. But, at the same time, influence happens when the child tries to change the thoughts, feelings or behaviors of the parent. Children constitute a large secondary market that influences the family purchase (McNeal, 1988). Research has concluded that children tend to influence more on the purchasing decisions of products that directly relate to them or that even affect them (Arzu, 2011; Sener, 2011; Foxman et al., 1989b; Atkin, 1978). Therefore, the following hypothesis is proposed:

H3.

The passive social power, expert and legitimate of preadolescent children have a direct positive effect on mother’s perception of the child’s influence (from child to mother) in toy stores.

On the other hand, it has been found that children have a lesser degree of influence in the decisions regarding products that have high costs and that are used by all the family (Foxman et al., 1989a). At the age of eight, most children become socialized consumers and enjoy having discretion to spend their own money (McNeal, 1992a, 1992b; Isler et al., 1987). Based on this, the fourth hypothesis arises:

H4.

The direct positive effect of the passive social power, expert and legitimate of preadolescent children on mother’s perception of the child’s influence (from child to mother) is moderate negatively for the purchase amount in the toy stores.

Methodology

The research undertaken has a non-experimental transversal correlational design. It is non-experimental given the fact that the unit is observed in its reality, meaning that the behaviors of the independent variables studied have already occurred, reason for which they have not been or could not have been manipulated. In addition, it is a transversal study because the data collection was done at a determined moment in time. Finally, it is correlational because it describes the relationship that exists between LSP and ESP types and the influence on the mothers’ purchasing behavior.

Sample

The target population for this study was mothers of Peruvian families who reside in Lima and go to purchase toys with their preadolescent children to major shopping centers.

The sampling method is probabilistic by clusters, where the units of analysis are encapsulated in various physical locations. Among all the possible clusters, toy stores of northern Lima shopping centers were chosen where the units of analysis were found – mothers with kids aged 8-11 years.

The National Institute for Statistics and information, Instituto Nacional de Estadística e Informática, indicates that the Peruvian population reached 31,151,643 inhabitants (INEI, 2015b). Considering that, on average, there are four members per household (Ipsos Perú, 2015), where Peruvian women have an average of 2.6 children (Andina, 2013; INEI, 2015a), and the population of preadolescent children between 6 and 11 years of age is 942,000 (INEI, 2015c).

Moreover, it was evidenced that the population of Metropolitan Lima is of 9,752,000 inhabitants, of which more than half of them live in East and North Lima (INEI, 2015c). North Lima is represented by 25.5 per cent of the total population of Metropolitan Lima, where the predominant socio-economic levels are C (39.6 per cent) and D (37.7 per cent), with San Martin de Porres and Comas being the districts with larger populations.

Because of the fact that this is a preliminary research, a pilot sample of 50 participants was chosen with the purpose of testing the research model.

Measurements

To elaborate the questionnaire, the scales of Swasy (1979) and from Beatty and Talpade (1994) were used as the basis (Appendix 1 and Appendix 2). These were adapted for the purpose and context of this research project (Appendix 3). The measuring instrument for social power is based on adaptations of Swasy (1979) of scales used to measure the perception of mothers according to the two types of passive social power selected: legitimate and expert. For the measurement of the mother’s perception of influence, an adaptation of Beatty and Talpade (1994) was used, which are two scales of relative influence; one based on the initiation stage and the other on the initial search/decision stage. All this is summed up in a questionnaire that consists of closed questions, which are the filter questions, with a Likert scale from 1 to 5.

The instruments for constructs and items used in this study have been previously validated by the authors of said concepts for social power and influence in purchasing. Initially, we recognized 85 from 150 items (French and Raven, 1959), thus we selected 31 validated items (Swasy, 1979), considering only the dimensions of ESP (8 items) and LSP (3 items). On the other hand, we recognized 9 validated items from 26 items (Beatty and Talpade, 1994) considering the dimensions of initial influence in purchasing – IIP (4 items) and decisional influence in purchasing – DIP (5 items) for the stages of influence in purchasing to get the final scale. The items, originally in English, were translated by a specialist and revised by three bilingual university professors in marketing field.

Passive social power (ESP and LSP), in accordance with Flurry and Burns (2005), is the degree to which a person is perceived to have the right to exert influence or the right that a person has of influencing on the behavior and/or beliefs of the other person. This variable will be measured through a basic Likert scale of five points, instrument elaborated and validated by Swasy (1979), which was adapted for the purpose of this study.

Influence perception in purchasing behavior (initial influence in purchasing – IIP – and decisional influence in purchasing – DIP). Influence perception is defined as the use of power to achieve a result (Coleman, 1973), in this situation, the purchase. In turn, by definition, the influence could only be achieved as a result of a process of exchange between two or more parts (Sprey, 1975). For this case, it will be the influence perception in mothers’ purchasing behavior, from mother to child. This variable will be measured over a frequency range of a five-point Likert instrument developed and validated by Beatty and Talpade (1994), which will be adapted for the study.

In addition, there is the variable of influence perception in mothers’ purchasing behavior, from child to mother, to purchase of a toy (INFL). For this variable, a Likert scale of 1-5 was used so that the results could be crossed with the other two scales previously described.

Control variables

Apart from the central variables to measure, various papers present other variables. The authors recommend considering and evaluating them to know the context upon which the central variables are given. These control variables are demographic variables: child age, child gender and expenditure level (purchase amount) per toy. This last control variable is used as a dummy variable used to measure moderation in two moments, lower or higher than S/.100.00 (dummy purchasing 100 – DP100), according to the average in spending from this sample.

Procedure

SPSS software was used to process the data. This way, a parametric analysis of the variables of social power with the influence in purchasing to establish the degree and direction (positive or negative) of the relationship was possible.

In the first place, to determine passive social power and influence in purchasing, a correlation analysis and a factor analysis was performed to reduce the number of variables, in addition to verifying that the items related to the described dimensions by the authors. That followed the stages of the procedure described by Hair et al. (2014).

So as to verify the degree of the relationship of the values of the variables, the Pearson coefficient of correlation was applied; this to done to determine the linear dependency between the two variables. The mathematical expression of the said coefficient takes into account the covariance and the variance of the two variables, which is expressed as follows:

ρxy=Cov(xy)Var(x) * Var(y)

The values that ρ_xy can have are in the range of [−1,1], so there is no correlation between the variables if ρ_xy = 0, while the correlation will be perfectly positive ρ_xy = 1 or perfectly negative if ρ_xy = −1.

For the analysis of passive social power, Table I shows that the correlations are significant among the items that correspond to ESP and LSP as independent factors. Both the studied dimensions of passive social power correlate according the bilateral significance with the majority of the items. Only the item LSP_1 has no correlation with the items LSP_2 and LSP_3.

Likewise, with respect to the influence in purchasing, Table II shows that the correlations are significant among the items that correspond to the factors of IIP and DIP. Only lower correlation indexes are reflected in IIP_1 and DIP_3 for their peers.

Subsequently, according to the factorial tests, the KMO and the Barlett tests, which is based on whether the contrasts among the partial correlations of the variables are sufficiently small, it can be said that a relationship among these variables exists. Having obtained a KMO index = 0.893 > 0.500 for passive social power and KMO index = 0.763 > 0.500 for influence in purchasing, the factorial analyses can be applied; and at the same time, the variables are grouped for data interpretation.

The Barlett’s sphericity test proves the null hypothesis that the correlations matrix is equal to the identity matrix. Then, being the identity matrix different to the correlations’ one, this allows the performance of factorial analysis. As the levels of sig. = 0.00 < 0.05, the factor analyses are possible.

Later, in accordance to the total variance resulting from the data, for the passive social power, two self-values were determined that explain 64.127 per cent of the total variance of the original data (Table III). As for the influence in purchasing, two self-values were determined that help explain 62.074 per cent of the total variance of the original data (Table IV).

Also, in the rotated component matrixes, Tables III and IV, the factors of ESP (8 items) and LSP (3 items) from a total of 11 items of passive social power, as well as the factors of IIP (4 items) and DIP (5 items) from a total of 9 items of influence in purchasing, were generated.

According to the results from the procedures and the use standards of factor loadings related to sample size needed (Hair et al., 2014), we used items with factor loading >0.65 and lower communality based on their dimension. For these reasons, we did not use the items LSP_1, IIP_1 and DIP_3.

Finally, the Cronbach’s α for each dimension was estimated, and the results are ESP > 0.913, LSP > 0.691, IIP > 0.840 and DIP > 0.878. Although LSP has a Cronbach’s α < 0.70, this is due to non-use of an item with lower factor loading and the only use two final items for the dimension. However, according to Nunally and Bernstein (1994), as a pilot test is acceptable.

Results

As per the results, Table V shows that the correlation analysis of the dimensions of ESP and LSP, both passive, present a correlation of 0.402 with sig. level <0.01, for which H1 is supported. This result supports the evidence of the positive relationship of the two passive social powers that the mother perceives with respect to her child in the context of toy stores in an emerging market. That is relevant, as these two passive social powers can be potentially causal on influence on purchasing behavior by mothers regarding toys.

Nevertheless, both variables present correlations under 0.5 with regard to the variables of influence in purchasing behavior (mother-child) with level of sig. >0.05; only ESP with IIP presents a correlation of 0.306 with a level of sig. <0.05. Thus, it demonstrates that there is no complete relationship between the said variables, deeming unnecessary the regression analysis that looks at the causality of said variables. Consequently, H2 is not supported.

On the other hand, both variables have correlations higher than 0.5 with regard to the variable of influence in purchasing behavior (child-mother) with levels of sig. <0.01; this demonstrates that there is a relationship between ESP and LSP variables with the influence in purchasing behavior from child to mother (INFL). The previous relationship, between ESP-LSP with INF, allows a regression test showed in the Table VI (Model 1). This technique has the purpose of determining the existence of causality between the mentioned variables.

A multiple linear regression was calculated to predict influence in mothers’ purchasing behavior (INFL), from child to mother, based on ESP and LSP. Looking at the regression analysis, it can be appreciated that the independent variables of ESP and LSP do affect the influence in mothers’ purchasing behavior from child to mother, having an R² = 0.438, and levels of sig. <0.01 for both independent variables as predictors of the dependent variable, for which H3 is supported.

After performing the regression analyses to test the moderation with respect to the ESP (Model 2a) and LSP (Model 2b) on influence in purchasing behavior from child to mother, in neither of the two cases the required level of sig. <0.05 in the regression coefficients is achieved; hence, H4 is not supported.

These results demonstrate that expenditure levels in purchases of products such as toys do not affect the relationship of passive social power, be it legitimate or expert, on influence in purchasing child-mother.

Discussion

Notwithstanding that the topic has been well addressed in the past, difficulties in finding specific recent sources about it were encountered. The topic has been left aside because the influence child-mother is assumed as being part of the direct reference group. It is recommended to test the whole model to more enrich the findings regarding the topic in question.

Having been this research project a pilot to test the research model in our context, it still remains a topic for future research in terms of sample size, location and other relevant variables that could affect the final results. To start with, the sample size needs to be increased to have a fair representation of Lima toy stores’ formal market and validate the results of the pilot test.

Then, application of the model should be expanded to other city areas to answer the question of do the results hold true for different toy store localizations? Any significant differences can then be researched. After this, other provinces in the country can be studied to compare results to the capital and amongst themselves to identify similarities and difference. The study could be replicated in other Latin American countries to see ramifications within this context.

This study focused on 8-11-year-old children, a variable that can have affected the preliminary results in a specific way that might not be representative of younger and older children. Hence, research could also include the application of this model to a younger group of children, as well as to teenage kids, to better understand the mothers’ perception of the relationship of their children influence with LSP and ESP types. Also, the application of the other three types of social power can be incorporated to the study. This to identify which one also applies in our local context and even identify if any one of them weighs more on the influence of toy purchases and how they compare to the ones already identified.

Being that the test subjects for this research were mothers, it would be interesting to survey by gender to ascertain if this would render any differences, and if so, to what level. Also, variables such as the children age, level of expenditure, marital status and the frequency of visits to toy stores could be included with more detail in future research projects.

Finally, the aspect of cultural differences and cultural contexts may very well arise as a result of expanding this study to other countries and regions, where the differences could be studied from the perspective of cultural differences or even the level of socio-economic development of the country or region: developed countries, emerging countries and underdeveloped countries.

Contribution

The information that this study brings can provide a better understanding of consumer behavior for toys, providing insights into motivations and the purchasing process of toys. This then allows companies to better identify their target markets and develop better marketing programs for their target markets by taking into consideration the role that children have on the purchasing decision process because of their levels of influence, as perceived by the mothers. Specifically, the information resulting from this research can help design more effective and efficient marketing communication programs to better communicate with their target markets, children and mothers, and specifically take advantage of the child-mother binomial.

For one, toy stores should focus more on the binomial child-mother and the manner in which influence emerges. Having the legitimate passive social power, more weight on the influence in purchasing means that there exists a possibility that the mother concedes to the child’s purchasing attempt, but this must be worked by the stores, specifically by the salespeople. Talking should be done to the binomial, to the mother and the child, for even though one pays, the other is who will be the user, and if the user does not want it, the purchase will not be finalized. In addition, the final user might have a level of expertise regarding the toy and be in the position to influence the mother’s purchasing behavior if she sees him or her as an expert on the subject.

Regarding part of the aspect store layout, the objective can be to make each toy store a space full of experiences for the children as the majority of mothers enter a toy store for the simple fact that their child will entertain himself/herself in the store, for which the following is proposed: provide spaces where children can play with the toys, allowing them to test toys, and the mother and the child tests and experiments with the toys. This even allows them, mother and children, to spend time together, providing mother with more insight into their children’s likes and preferences.

Conclusions

The passive ESP does not positively correlate to any of the dimensions of the influence in purchasing (mother-child). The passive LSP does not positively correlate to any of the dimensions of the influence in purchasing (mother-child). In conclusion, the results of Flurry and Burns (2005) are not supported in this international context.

Additionally, taking as external dependent variable the influence in purchasing (child-mother) of a toy, the result of crossing it with the four dimensions, this variable correlated principally with the passive social power, ESP and LSP. Thus, we can say that the passive social power has the two independent dimensions significant for the present study.

Specifically, within the construct of passive social power, the dimension with the best correlation is the passive LSP. Basing ourselves on the theory (Flurry and Burns, 2005; Swasy, 1979; French and Raven, 1959), it is an innate power that is found in the person that exerts the power because the person only has in common with the other person a strong link, in this case of child-mother.

Figures

Figure AI.

Table I.

Pearson correlation – variables of passive social power

N = 67	ESP_1	ESP_2	ESP_3	ESP_4	ESP_5	ESP_6	ESP_7	ESP_8	LSP_1	LSP_2	LSP_3
ESP_1	1
ESP_2	0.428**	1
ESP_3	0.506**	0.595**	1
ESP_4	0.527**	0.548**	0.818**	1
ESP_5	0.504**	0.754**	0.560**	0.577**	1
ESP_6	0.671**	0.567**	0.615**	0.679**	0.703**	1
ESP_7	0.598**	0.456**	0.545**	0.581**	0.592**	0.679**	1
ESP_8	0.560**	0.401**	0.379**	0.545**	0.478**	0.527**	0.443**	1
LSP_1	0.527**	0.455**	0.396**	0.551**	0.383**	0.507**	0.374**	0.561**	1
LSP_2	0.466**	0.313**	0.387**	0.388**	0.495**	0.459**	0.419**	0.302*	0.232	1
LSP_3	0.156	0.104	0.144	0.037	0.194	0.094	0.206	0.114	−0.005	0.535**	1
Mean	3.597	3.358	3.328	3.373	3.284	3.537	3.015	4.224	3.925	2.881	2.134
SD	1.016	1.069	1.147	1.191	1.056	1.078	1.108	1.012	0.990	1.080	0.919

Notes:

**Correlation is significant at the 0.01 level (two-tailed).

*Correlation is significant at the 0.05 level (two-tailed); authors’ tabulation based on SPSS output

Table II.

Pearson correlation – variables of influence in purchasing

N = 67	IIP_1	IIP_2	IIP_3	IIP_4	DIP_1	DIP_2	DIP_3	DIP_4	DIP_5
IIP_1	1
IIP_2	0.319**	1
IIP_3	0.281*	0.620**	1
IIP_4	0.270*	0.874**	0.621**	1
DIP_1	0.291*	0.429**	0.380**	0.426**	1
DIP_2	0.289*	0.416**	0.349**	0.377**	0.684**	1
DIP_3	0.131	0.215	0.319**	0.284*	0.245*	0.374**	1
DIP_4	0.108	0.234	0.108	0.259*	0.457**	0.567**	0.370**	1
DIP_5	0.313**	0.255*	0.245*	0.214	0.504**	0.545**	0.301*	0.684**	1
Mean	3.030	3.418	3.433	3.328	3.761	3.791	3.403	3.567	3.672
SD	0.778	1.047	1.048	1.106	1.102	1.038	0.922	0.874	0.975

Notes:

**

Correlation is significant at the 0.01 level (two-tailed);

*

Correlation is significant at the 0.05 level (two-tailed); authors’ tabulation based on SPSS output

Table III.

Rotated component matrix^a of passive social power

N = 67	Component
	ESPb	LSPb	Communality
ESP_1	0.701	0.287	0.574
ESP_2	0.736	0.124	0.557
ESP_3	0.798	0.143	0.657
ESP_4	0.881	0.040	0.778
ESP_5	0.738	0.347	0.665
ESP_6	0.829	0.231	0.740
ESP_7	0.696	0.342	0.601
ESP_8	0.677	0.083	0.465
LSP_1	0.702	−0.098	0.502
LSP_2	0.344	0.830	0.808
LSP_3	−0.044	0.840	0.707
Eigenvalue	5.232	1.822
% of variance explained	47.566	16.561	64.127
Cronbach’s α	0.913	0.691

Notes:

Extraction method: principal component analysis;

^a

rotation converged in three iterations;

^b

italic data are items that represent the component in order to the defined construct; authors’ tabulation based on SPSS output

Table IV.

Rotated component matrix^a of influence in purchasing

N = 67	Component
	IIPb	DIPb	Communality
IIP_1	0.279	0.322	0.182
IIP_2	0.185	0.908	0.859
IIP_3	0.159	0.799	0.664
IIP_4	0.168	0.917	0.870
DIP_1	0.735	0.360	0.669
DIP_2	0.809	0.294	0.741
DIP_3	0.444	0.238	0.254
DIP_4	0.810	0.035	0.657
DIP_5	0.828	0.075	0.691
Eigenvalue	2.899	2.687
% of variance explained	32.215	29.859	62.074
Cronbach’s α	0.840	0.878

Notes:

Extraction method: principal component analysis;

^a

rotation converged in three iterations;

^b

italic data are items that represent the component in order to the defined construct; authors’ tabulation based on SPSS output

Table V.

Pearson correlation of dimensions

N = 67	INFL	IIP	DIP	ESP	LSP
INFL	1
IIP	0.221	1
DIP	0.172	0.426**	1
ESP	0.505**	0.306*	0.089	1
LSP	0.612**	0.063	−0.034	0.402**	1
Mean	3.388	3.393	3.698	3.465	2.507
SD	0.870	0.957	0.823	0.856	0.877

Notes:

**

Correlation is significant at the 0.01 level (two-tailed);

*

Correlation is significant at the 0.05 level (two-tailed); authors’ tabulation based on SPSS output

Table VI.

Model estimates: regression coefficients (t-values)

	INFLa
	Model 1	VIFb	Model 2a	VIFb	Model 2b	VIFb
ESP	0.314 (3.064)**	1.192	0.492 (3.360)**	1.772
LSP	0.484 (4.842)**	1.192			0.651 (4.977)**	1.765
DP100			−0.336 (−0.419)	16.621	0.150 (0.283)	8.705
ESP × DP100			0.060 (0.268)	17.879
LSP × DP100					−0.097 (−0.490)	9.614
Constant	1.087 (3.092)**		1.725 (3.355)**		1.790 (5.231)**
Adjusted R2	0.438		0.225		0.350
Durbin-Watsonc	2.025		1.957		1.972
F-statistic	26.679		7.402		12.858
N	67		67		67

Notes:

^a

Dependent variable: influence in purchasing (child-mother);

^b

VIF > 5 multicollinearity problems are potentially severe;

^c

Durbin–Watson statistic no reject H₀: autocorrelation does not exist;

**

p < 0.01; authors’ tabulation based on SPSS output

Table AI.

Social power

Item	Reward	Loading	Alpha*
1	If I do not comply with A, I will not be rewarded	0.67	–
2	The only reason for doing as A suggests is to obtain good things in return	0.67	0.67
3	I want to do as A suggests only because of the good things A will give me for complying	0.66	0.74
4	A has the ability to reward me (in some manner) if I do as A suggests	0.65	0.76
5	If I do not do as A suggests I will not receive good things from A	0.63	0.81
6	In this situation I am dependent on A’s willingness to grant me good things	0.60	0.82
	Reference
1	In general, A’s opinions and values are similar to mine	0.77	–
2	Being similar to A is good	0.68	0.71
3	I want to be similar to A	0.57	0.80
4	In this situation my attitudes are similar to A’s	0.65	0.79
5	I would like to act very similar to the way A would act in this situation	0.63	0.81
6	In this situation my behavior is similar to A’s	0.62	0.83
	Information
1	The information provided by A about this situation makes sense	0.67	–
2	The information A provided is logical	0.62	0.67
3	I will seriously consider A’s request because it is based on good reasoning	0.61	0.74
	Coercion
1	A can harm me in some manner if I do not do as A suggests	0.77	–
2	If I do not do as A suggests, A will punish me	0.76	0.77
3	Something bad will happen to me if I don’t do as A requests and A finds out	0.72	0.81
4	I had better do as A suggests in order to prevent something bad from happening to me	0.65	0.82
5	A might do something which is unpleasant to those who do not do as A suggests	0.64	0.84
	Expertise
1	I trust A’s judgment	0.74	–
2	A’s expertise makes him/her more likely to be right	0.73	0.63
3	A has a lot of experience and usually knows best	0.70	0.74
4	A knows best in this situation	0.69	0.80
5	A’s knowledge usually makes him/her right	0.66	0.83
6	I trust A’s judgment in this situation	0.65	0.85
7	In this situation I don’t know as much about what should be done as A does	0.61	0.85
8	A is intelligent	0.60	0.86
	Legitimate
1	It is my duty to comply with A	0.66	–
2	Because of A’s position he has the right to influence my behavior	0.61	0.38
3	I am obligated to do as A suggests	0.60	0.59

Source: Swasy (1979)

Table AII.

Influence perception in purchasing behavior

Initiation Stage
1	Bringing up the idea to buy the product
2	Getting people to realize that this product was needed
3	Realizing that this product would be useful to have
4	Getting others to start thinking about buying the product
Search/Decision Stage
1	Visiting the store(s) to look for different brands/models of the product
2	Examining different brands/models at the store
3	Picking up the product from the store
4	Deciding on the brand/model that was finally purchased
5	Deciding on which store to actually buy the product from

Source: Beatty and Talpade (1994)