Discussion

csht

export-behavior-small-medium-sized-manufacturers-minnesota-quant-corr.pdf

Home >Psychology homework help >Discussion

Study of the Export Behavior of Small and

Medium-Sized Manufacturers in Minnesota

Using Quantitative Correlational Analysis

Contributors: R. Don Keysser

Pub. Date: 2020

Product: SAGE Research Methods Cases

Methods: Correlation, Survey research, Research questions

Disciplines: Business and Management

Access Date: January 30, 2020

Academic Level: Postgraduate

Publishing Company: SAGE Publications Ltd

City: London

Online ISBN: 9781529713978

DOI: https://dx.doi.org/10.4135/9781529713978

This PDF has been generated from SAGE Research Methods Cases.

https://dx.doi.org/10.4135/9781529713978

Abstract

This case study describes my use of quantitative correlational analysis in analyzing the exporting behavior

of small and medium-sized businesses (SMBs), for my doctoral research. The data were generated through

an online survey conducted with 375 members of a trade group in Minnesota, the United States (the

Manufacturers Alliance); the trade group consists primarily of small and medium-sized family-owned

manufacturing firms. The goal was to understand the correlation of various company characteristics with

exporting behavior. Quantitative indices of the “Propensity to Export” and the “Intensity of Exporting” were

developed and then correlated with five characteristics that are theorized to affect the export development

process of SMBs: the nature of the management structure of the SMBs, the size of the SMBs, the location of

the SMB (rural vs. urban), the industry cluster of the SMBs, and the access SMBs perceive they have to the

external resources they need for export development. The results were used to develop recommendations

on the development of exporting by manufacturing SMBs and on the opportunities for stakeholders who

work with SMBs. This case reviews the process of conducting research using an Internet-based survey

and quantitative correlational analysis, provides several factors that can contribute to the success of this

approach, and discusses the relative merits of a quantitative versus qualitative method of research.

Learning Outcomes

By the end of this case, students should be able to

• Develop an understanding of some of the technical challenges when conducting quantitative

research in support of an advanced degree, including a doctoral dissertation

• Understand the mechanics and processes of conducting a quantitative research process, using an

email-based survey tool and mathematical statistical analysis tools

• Understand the distinction between causality and correlation, and how to structure a quantitative

analysis to focus on correlation

• Understand the difference between a quantitative analysis and a qualitative analysis

Introduction

My study into the export development process and exporting behavior of small and medium-sized businesses

(SMBs) began with the following question: Why do U.S. SMBs export at a relatively low rate, when compared

with larger U.S. corporations, and even when compared with non-U.S. SMBs? The estimate of the number

of U.S. SMBs that engage in some level of exporting ranges from 5% to 15%, depending on the survey, the

percentage of exporting sales considered significant, and the sector of the economy (EIM, 2010). Given the

potential benefits to exporting, and the strong emphasis on promoting exporting by SMBs at the governmental

SAGE

SAGE Research Methods Cases

Page 2 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

level, the question then becomes what factors, or company characteristics, might be related to the exporting

behavior of U.S. SMBs, which (a) represent over 90% of employer firms; (b) generate over 64% of net new

jobs in private sector; (c) represent over 40% of total private-sector payroll; (d) provide 46% of private-sector

output; and (e) create the majority of growth and innovation (Dennis, 2004).

As a consultant to manufacturing SMBs, I regularly encounter SMBs that do not engage in a significant level

of exporting, whether out of disinterest, an exclusive focus on domestic markets, or a fear of the unknown

complexities inherent in initiating exporting. While there are certainly valid reasons for hesitating to engage

actively in exporting, there is considerable evidence in the literature, both academic and business, of the

advantages that can accrue to an SMB that does engage in exporting (McCracken, 2013; Soroka, 2011).

These advantages include an increase in enterprise value; expanded market opportunities and exposure

to a new customer base; improved cash flow and increased revenues; mitigation of single-market risk, as

exporters diversify into other economies and countries; a gain in knowledge of new technologies and ideas;

decrease in production costs through greater economies of scale and improved global supply chain; and

extension of product life cycle, and increased net margins.

None of these outcomes are assured or panaceas; exporting is a complex and risky enterprise. However,

over 90% of the world’s population and 70% of the world’s purchasing power lie outside the United States,

and several recent studies (Baily, 2012; Daud, 2013; Freund, 2014; Gootman, Slo, Shenkar, & Stewart, 2014;

McCracken, 2013) have shown an increased interest by U.S. SMBs in exporting, suggesting the importance

of identifying how SMBs can be assisted in increasing their level of exporting. Business consultants regularly

advise their manufacturing SMB clients to give serious consideration to exporting, and to defining the internal

and external resources needed for such a strategic choice.

Project Overview and Context

My study was not intended to answer the question of “why don’t SMBs export at a higher rate?” That answer

would suggest causality, and a different type of analysis and mathematics. Causality would be measured by

such instruments as regression analysis, and the use of control groups, whereby the researcher can show

empirically and through data that “A” causes “B,” especially in a temporal sense. Given all of the complexities

in measuring cause-and-effect in a field as complex and multi-tiered as international exporting, it is likely not

a sound exercise to attempt to prove causality.

Instead, my study focused on correlations between sets of characteristics. Its purpose was to examine the

correlation between the levels of interest SMBs have in exporting (propensity), and their actual engagement

in exporting (intensity), controlling for specific company characteristics, to assist in understanding the factors

affecting export development for Minnesota manufacturing SMBs.

Research Design

SAGE

SAGE Research Methods Cases

Page 3 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

The first step in developing my research program was to make a decision about conducting a quantitative

or a qualitative analysis. These two choices are not the stark opposites they may seem, but should instead

be seen as two points on a continuum. Studies tend to be more quantitative or more qualitative, rather than

absolute. One distinction between the two is in the choice of expression: quantitative analyses are typically

expressed in numerical terms while qualitative analyses are expressed in words. A quantitative analysis,

as I chose here, is “a means of testing objective theories by examining the relationships among variables”

(Creswell, 2009, p. 4), where those variables are expressed in numerical form. In an ideal world, one in

which I had the time to do an extensive research project, I would have preferred to conduct a mixed-methods

research, in which the qualitative analysis would have added more subjective “color commentary” and depth

to the quantitative analysis.

Once I selected a quantitative approach, my study then began with the creation of two quantitative indices:

“Propensity to Export” (interest in exporting) and “Intensity of Exporting” (actual level of exporting), creating

a 2 × 2 matrix of companies based on these two indices, which were then measured among the Minnesota

manufacturing SMBs that participated in an online survey, and the relationship between those two variables,

and to determine the extent to which such a relationship is changed by specific company characteristics: type

of management structure, size, location, industrial cluster, and access to outside resources. In this way, the

study was intended to consider factors that may be associated (correlated) with levels of exporting intensity,

as an explanation, rather than “causing” export behavior. These indices were developed partly from my

literature review of the prevailing knowledge and research that has already been conducted on SMB export

development.

Method in Action

I began this research by making several methodological decisions. First, I focused on a target with which I

am familiar, and for which there are numerous candidates: SMBs, primarily family-owned, in Minnesota. This

is a group with which I, as a business finance consultant, work on a regular basis. Second, as discussed

earlier, I did not seek to establish causality, to answer the question of “why,” but rather focused on correlations

between company characteristics and patterns of exporting behavior, to answer the question “what.” While it

may be tempting to further imply causality from the correlations, and a deep understanding of the dynamics

of small-firm exporting may lead one in that direction intuitively, it is very important to make a clear distinction

in the research findings between causality and correlation.

Third, I chose to employ a quantitative analysis, since I had access to a very robust database, rather than

attempt to do a qualitative analysis through a series of one-on-one interviews. Perhaps I am more comfortable

dealing with the perceived precision of numbers than I am with the more subjective nature of interviews,

something that any researcher needs to understand about himself or herself. Part of the reason for my focus

on a quantitative analysis of survey data is that I had the very strong support of a trade association; more

on that later. Fourth, I decided to use an online survey tool because of its ease of use, low cost, and relative

SAGE

SAGE Research Methods Cases

Page 4 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

speed, and because the data generated by this survey tool were easily accommodated in my statistical tools:

Statistical Package for the Social Sciences (SPSS) and Microsoft Excel.

Fifth, to accommodate a quantitative analysis and obtain a high return rate, the survey was limited to only

11 questions, all but one of which were structured as closed-ended multiple choice questions, and some

structured on a Likert-type scale, resulting in nominal or ordinal scale data. It was also my understanding

(as well as that of my dissertation committee) that anything much larger than 10 questions runs the risk of

fewer responses. Sixth, based on my understanding of the dynamics of small firm management and on my

literature review, I selected a series of company characteristics that I thought could have a logical relationship

to exporting behavior, including questions on company management structure, size, industrial sector, and

location. For example, I have repeatedly observed, and the literature on the subject supports this perspective,

that family-owned businesses, when actively managed by family members, respond differently to situations of

risk and new market challenges than do family-owned businesses where senior management is from outside

the family.

For this study, I used an online survey, through the SurveyMonkey platform, distributed through email to all

375 members of a local trade association, the Manufacturers Alliance (MA), which consists almost exclusively

of family-owned Minnesota-based manufacturing SMBs. In most instances, the survey was sent directly to the

CEO of these companies. I choose SurveyMonkey because of its ease of use in structuring the survey and

in responding to it online, its strong data tracking capabilities, its ability to export data directly into statistical

software, and its low cost.

There were 79 responses, for a response rate of 21.1%, in part due to the strong active support by the MA of

this research. For an online survey through email, particularly when I had no immediate relationship with the

individuals in the database (other than the sponsorship of the trade association, as discussed later), this is a

relatively high response rate. A response rate of 10% would have been more the norm. It is to the researcher’s

advantage, however, to do whatever is possible to obtain as high a return rate as feasible, as that adds to the

validity of the research and the conclusions he or she draws from it.

I analyzed the data using SPSS, for which I purchased a 6-month license. As I later came to understand,

I could have just as easily performed these analyses using the Statistical Add-in package for Excel, which

is free. However, SPSS does have the advantage of making it easier to import survey data directly into the

software.

First, I generated a series of cross-tabulations on the bivariate distribution of the participant population by

the two indices I had created, Intensity of Exporting and Propensity to Export, and by the previously selected

company characteristics. Cross-tabulation analysis uses data tables that present the results of an entire

population of respondents, as well as results from sub-groups within that population, allowing the researcher

to examine relationships within the data among the respondents. It can be used to quantitatively analyze the

relationships between multiple variables and sets of variables for levels of correlation.

SAGE

SAGE Research Methods Cases

Page 5 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

Second, I calculated a series of Pearson’s correlation coefficients (PCCs), both for the relationship between

Propensity and Intensity on their own, and then for this relationship controlled by the series of five company

characteristics identified above. PCC is a technique for investigating the relationships between two

quantitative scaled variables, and a measure of the strength of association between those two variables.

To accommodate the cross-tabulation and Pearson analyses, the questions in the survey instrument were

structured on an ordinal or nominal scale, in some instances using a Likert-type scale. This is a tool that poses

a subjective question, but then gives the respondent a choice of 5 to 10 responses, all scaled, to provide

some measure of ordinality.

As an example of a question posed on a Likert-type scale, one of the questions asked was, “What is your

company’s level of interest in engaging in exporting?” The possible answers were (a) not engaged, not

interested; (b) not engaged, interested in pursuing it; (c) already engaged, not interested in expanding; (d)

already engaged, interested in further expanding; and (e) already actively engaged, fully committed. By

scaling the responses in these five steps of increasing level of interest, the responses can then be correlated

against other questions. For example, a non-scaled question asked was, “Where is your primary Minnesota

facility?” The possible responses were (a) Twin Cities metro area (Minneapolis/St. Paul) and (b) out-state

Minnesota (non-metro). A correlation between these two sets of responses can lead to an understand of

whether or not there is a correlation between a company’s location (urban vs. out-state) and Propensity to

Export.

A critical element to this research’s success was the strong active support of the survey sponsor, the MA, who

made its entire database available to me. The MA sent out a notice through email under its letterhead and

signature and through its newsletter about the forthcoming survey, urging full cooperation from their members.

A follow-up email was sent out about 5 days later, again urging a quick response to the survey. This level of

active support from a trusted professional trade group contributed to a 21% response rate.

Summary of Frequency Distributions and Cross-Tabulation

Distributions

Two indices were created, using the questions and responses from this survey: the Propensity to Export

and the Intensity of Exporting. “Propensity” explored the interest a respondent company had in engaging

in exporting (whether or not it actually did export), and “Intensity” measured the actual level of exporting in

which the company engaged, aside from its level of interest. Several of the survey questions were directed

at determining the level of interest by companies in exporting, as well as actual levels of exporting. One of

the areas about which I was curious was the extent of “accidental exporting,” something discussed in the

literature, in which a company has a very robust e-commerce website and picks up international business

without any overt effort at marketing overseas, such as hiring sales representatives and attending trade

shows.

SAGE

SAGE Research Methods Cases

Page 6 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

Some of the results from this analysis are summarized here. When measuring company size against both

Propensity and Intensity, a modest correlation emerged that the larger SMBs tended to have a higher

intensity of exporting and a higher propensity to export. Approximately half of the participating SMBs reported

a management structure that was either mostly or 100% outside management, and there was a modest

association between that management structure and a higher intensity. The pattern between management

structure and propensity was less clear; if anything, there seemed to be a weak relationship between

propensity and the family-management structure, while there was a stronger relationship between intensity

and the outside management form of management structure. It is possible that family-managed SMBs have

an interest in exporting but have failed to follow-up on this interest, while outside-managed SMBs have taken

more of the steps needed to initiate exporting and are therefore starting to record international sales.

There was a similar pattern for both Propensity and Intensity when paired with the use of resources. The

majority of the SMBs never or rarely used the governmental resources available to them, but made fairly

regular use of the private resources of local financial institutions and private service providers. A similar

pattern for both Propensity and Intensity held for the perceived access to resources, as did for the use

of resources. Almost all of the respondents did not view the public resources as being accessible, but a

significant percentage of them found the private resources to be accessible.

The discussion on the association between Level of Interest and other characteristics showed some

interesting patterns. There was a clear association between the level of interest and management structure,

where outside-managed firms (and balanced/mixed management firms) presented a stronger level of interest

in exporting than the family-managed firms. The participating SMBs in the general manufacturing category

indicated a higher level of interest, relative to their total sample size, compared with the medical device SMBs.

There was also an association between level of interest and perceived access to resources, with a higher

percentage of participating SMBs showing a link between an interest in exporting and relatively accessible

resources.

Summary of Correlational Analysis

The essence of this analysis was to determine the correlations between company characteristics (e.g., extent

of family management) and both Propensity and Intensity. There is a fairly extensive literature on the general

topic of small-business export development, although most of it is focused on emerging markets rather than

established Western markets.

Several issues in the literature were supported by the results of the study. First, the reasonably strong

correlation between Propensity and Intensity suggests that there is a higher level of interest of SMBs in

exporting than in the actual level of engagement in exporting. The correlation would suggest that a higher

level of interest should result in a higher level of activity. Yet, it is reasonable to conclude from the data that

there are more SMBs interested in exporting than are engaged in exporting. It is not apparent whether these

data mean that there are SMBs who are interested in exporting but have not yet made the decision to initiate

SAGE

SAGE Research Methods Cases

Page 7 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

exporting, or SMBs that are interested and have started exporting, but so far without success (since intensity

only measures successful results, not efforts). The question then becomes what company characteristics are

correlated with engagement in exporting.

Second, the type of the management structure seemed to influence exporting behavior, in that the correlation

between Propensity and Intensity varied by form of management structure. As suggested in the literature,

SMBs that bring in at least some outside management had stronger correlations between Propensity and

Intensity than SMBs that had 100% family management, a finding that also emerged from the cross-

tabulations. There appears to be an advantage, in terms of facilitating exporting behavior, in bringing in

outside management, in terms of new skill sets and experience, and a heightened willingness to accept the

risks of exporting.

Third, there was a relationship between the perceived access to resources, and the correlation between

propensity and intensity. SMBs seemed to engage in exporting to the extent that they perceived a reasonable

level of access to those external technical resources needed to engage in exporting successfully. At the

same time, it was interesting to note how few of these resources are actually used by SMBs. Governmental

resources, including the Small Business Administration (SBA), the Export-Import Bank, the Trade Office, and

the U.S. Commercial Services, are virtually unused by SMBs. The only resources used by SMBs on a regular

basis were the private resources: local financial institutions, for capital, and professional service providers, for

technical assistance.

Finally, there is a moderate correlation between Propensity and Intensity with SMBs in the medical device

cluster, suggesting that these firms, being very high-tech and engaged in an industry that is global by its

nature, are successfully engaged in exporting. However, the cross-tabulation analysis showed a slightly

higher level of interest in exporting among general manufacturing SMBs.

The importance of this analysis, from my perspective, was threefold. First, it provides valuable insights to

governmental agencies charged with the promotion of small-business exporting (e.g., the SBA, the ExIm

Bank) on what some of the elements of a successful export development process should include. Second,

it can provide useful information, for example, on the availability of government resources, to SMBs that are

considering engaging in exporting. Third, it becomes a valuable tool for the bankers and consultants who

serve the SMB market, in providing insights on how best to assist their clients.

Primary Success Factors in Conducting This Quantitative Research

There were several elements of my dissertation process that I think contributed to the success of the effort.

Writing a dissertation is a grueling process, taking many months (in my case, about 12 months). Now that I

am serving on two doctoral dissertation committees myself, I am seeing both the perils of a poorly planned

process and the rewards of a well-planned process.

SAGE

SAGE Research Methods Cases

Page 8 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

Engage a Supportive Sponsor

A critical element for the success of this study was the active support of my sponsor, the MA. The MA’s

support included (a) providing its entire database; (b) emailing its membership to alert them that the survey

was coming via email and was important to the MA; (c) sending out the survey under its name and letterhead;

and (d) sending a follow-up email 5 days later reminding the recipients of the importance of the email. In

return, I provided the MA, and other economic development agencies in the state, with a summary of my final

report. I can’t overestimate the value of this strong active support from a sponsor; it is hard to imagine as high

a response rate (21%) as I received without their support.

I was able to develop the MA’s support because I was already a member of the trade group, as a service

provider, and as a trainer for some of their programs and as a host and sponsor of their events. I also built a

very good relationship with their senior program manager, whose enthusiasm for this project was invaluable.

To find a similar sponsor, I suggest that a researcher become very active in whatever professional community

in which he or she is involved.

Define Your Study Question Carefully

Create a narrow and specific research question that lends itself to a clear definition and to straight-forward

answers for which there is good data and that can be quantified and measured. For this study, the research

question was as follows: What is the relationship between the propensity to export and the intensity of

exporting of Minnesota manufacturing SMBs, based on specific company characteristics (management

structure, perceived access to resources, risk aversion, size, and industry)? Each of these variables is defined

in the study and is inherently measurable on a quantitative scale, which then makes a cross-tabulation and

PCC study feasible.

Use a Short Online Survey

I chose to use the SurveyMonkey platform for its ease of use and simplicity. The survey itself was short,

written in non-technical language, and easy to understand: 10 questions, structured on a 4-point Likert-type

scale, plus one open-ended question. I used the Likert-type scale technique because it allows a nuanced

scaled response rather than a simple yes/no or true/false response. Yet, it results in data that are structured

on a nominal or ordinal scale, permitting quantitative analysis.

My respondents were CEOs of manufacturing firms and would likely not be receptive to a lengthy survey. I

used the open-ended question to analyze the frequency with which specific words were used. For example,

the words “capital” and “funds” came up frequently, suggesting that for many would-be exporting companies,

the perceived lack of capital is a major obstacle. As an incentive to the recipients, I offered them a free 2-hr

consulting session on international trade; to me, this seemed a more meaningful inducement than a USD 50

prepaid credit card. Several of the respondents took me up on my offer, which had the side benefit of giving

SAGE

SAGE Research Methods Cases

Page 9 of 12 Study of the Export Behavior of Small and Medium-Sized Manufacturers in

Minnesota Using Quantitative Correlational Analysis

me a new client.

Use Appropriate Statistical Techniques

It is important to distinguish between causality and correlation, as discussed earlier. We naturally tend to

think of phenomena as being causally related, but that is a difficult analysis, and too often makes incorrect

assumptions about the phenomena being studied. Correlation research, in comparison, determines the

degree or extent to which a relationship exists between two or more variables, without implying a causal

relationship. It is one thing to say that “company size is correlated with exporting behavior,” and quite a

different thing, and far more difficult to prove, to say that “company size causes export behavior.” Another term

for correlational research is associational research—when the relationships between two or more variables

are analyzed without any attempt to influence or control them.

Link the Research Question With the Literature Review

I began my literature review very broadly, without a predefined research question, and the question I ended

up with flowed naturally from my readings. I came across specific findings from earlier research that I decided

to test within the specific population I had available. This explicit link between the literature review and the

research question is essential for a meaningful and viable research study. As it turned out, the findings

from this study strongly supported several of the theories in the current literature, including the correlation

between company family ownership/management and exporting behavior. However, some of my findings did

not support the literature, including the possible correlation between company location (metro vs. non-metro)

and Propensity to Export.

Exercises and Discussion Questions

1. Considering the research question posed in this study, discuss the merits of using the

quantitative correlational approach. Could this study have been structured as a qualitative or

mixed-methods study? Explain your answer.

2. Discuss the implications of using different approaches for a quantitative analysis: for example,

a different survey platform, more or fewer questions.

3. What is the distinction between correlation and causality? Explain the advantages and

disadvantages of approaching your research question from each perspective.

4. Considering your own research or a research topic of your choice, how would you structure

the quantitative correlation analysis? What indices would you create? Explain your answer.