Mobile Edition
Print
Get the Mag
Weekly UpdatesAbstract
During 2007, a needs assessment of the Oregon forest products industry was conducted. Two hundred and eighty-eight responses were compared to 441 responses from a 1995 survey of the educational needs of Oregon forest products firms. Identifying New Markets ranked #1 for both surveys, while Product Pricing ranked #4 in 1995 and #2 in 2007 and Strategic Market Planning ranked #11 in 1995 and #3 in 2007. For those companies with 20 to 100 employees, however, Competitive Positioning and Lean Manufacturing ranked #1 and #2, respectively. For companies with more than 100 employees, Environmental Protection Agency, Department of Environmental Quality Regulations and Safety Regulations ranked #1 and #2, respectively.
**********
It had been 12 years since a comprehensive educational needs assessment of the Oregon wood products industry was completed (Hansen and Smith 1997). Much has changed during this time. Several noteworthy examples include the creation and collapse of a secondary wood products industry association, the takeover of Willamette Industries by Weyerhaeuser Company, closures of additional Oregon mills, primarily sawmills and plywood plants, and the continuation of the U.S. forest products industry divesting its timberlands from its product businesses and selling them to timberland investment management organizations (TIMOS) and real estate investment trusts (REITS).
From 1990 through 2007, Oregon wood products employment declined by almost 36 percent. There was a 28 percent decline for sawmills and wood preservation, about 44 percent for plywood and engineered wood products, and about 33 percent for other wood product employers. During the recession years, 2001 to 2003, the wood products sector lost approximately 4,000 (11%) jobs (Myers 2008). Some of the lost employment may be explained by improved efficiencies as Oregon still ranked #1 among all states in lumber production (WWPA 2007) and plywood production (APA 2008).
Non-response bias
Since we are interested in using the data to design educational programs for the Oregon forest products industry, it is important that we can say, with some certainty, that the returned data are representative of the needs of the industry, even those of the non-response companies. It has been reported for surveys with low return rates, not uncommon for mail and internet surveys such as this one, that non-response bias may invalidate generalizing the response data across the entire sample population (Armstrong and Overton 1977). If respondents differ significantly from non-respondents then it would not be valid to assume that the data received from the survey directly relates to the entire sample population. For this reason, it is important to test for non-response bias before generalizing those results. One method of testing for non-response bias assumes that participants who respond less readily, such as those that respond only after increased stimulus such as a second mailing, are expected to be more like non-respondents. If they are found to be similar to the first respondents, then generalizing the response data across the sample population should be valid (Armstrong and Overton 1977, Hansen et al. 2008).
Comparison with past surveys
The results of this study were compared with the results of similar Oregon studies: one completed in 1988 (Brown and Niemiec 1997) and one completed in 1995. The 1988 survey only targeted sawmills. The information requested in this study was very similar to information requested in the 1995 study except that the 1995 study included information for both Oregon and Virgina (Hansen and Smith 1997). Hansen and Smith (1997) reported mailing to 1,286 Oregon firms for the 1995 survey.
Methods
Sampling
A questionnaire, nearly identical to one assessing the needs of Oregon's forest products industry in 1995 (Hansen and Smith 1997), was mailed to 1,532 firms in Oregon. The list of firms was composed of data collected by the Oregon Wood Innovation Center at Oregon State University and consisted of the most inclusive list known to the authors for identifying forest products firms operating in the state.
Questionnaire design
The design of the questionnaire was based on the discrepancy method described by Bratkovich and Miller (1993). A discrepancy need is described as the difference between what is and what should be. Learners will often respond that they need more training or help in subjects that are very important to them, but also in subjects that they currently possess a great deal of information, knowledge, and skills. In the discrepancy method, participants' information, knowledge, and skill level are assessed. A hierarchal list of needs is built by combining what respondents feel are important with what they also respond to as not having enough information, knowledge, or skill. The perceived Educational Need was calculated as: (Importance Rating--Current Knowledge Rating) X Mean Importance Rating (Bratkovich and Miller 1993). The discrepancy method was also used to design the 1995 survey for Oregon and Virginia forest products firms (Hansen and Smith 1997).
The questionnaire was designed to assess 36 subject areas in terms of importance and knowledge (Tables 1 and 2). The difference between the 1995 and 2007 questionnaires was the addition of three topic areas to the 2007 questionnaire: lean manufacturing, customer relationship management, and biomass/biofuels. Demographic data collected from the survey included primary product line and company size based on the number of employees.
Survey administration
An initial cover letter, questionnaire, and return envelope were sent to 1,532 firms in Oregon. The mailing list was compiled from a comprehensive database of primary and secondary forest products manufacturers in Oregon. For this study, participants were given two options to complete the questionnaire. They could return it to Oregon State University in the enclosed self-addressed metered envelope or they could access a website and complete the questionnaire online. After approximately 4 weeks, non-respondents were sent a second identical questionnaire along with the original cover letter and a second letter explaining that this was the second and final request. The first mailing allowed participants to opt out of the survey by returning the questionnaire unanswered. Respondents were asked to rate the importance and knowledge of company personnel in a subject, based on a 5-point scale (1 = level of lowest importance and 5 = level of highest importance).
Non-response bias was tested using the method described by Armstrong and Overton (1977). The first 30 respondents were compared to the last 30 respondents by each category of firm size, that is, the number of employees. Independent t-tests were used to compare differences in means of the two subsamples (Hansen et al. 2008).
Results and discussion
One-hundred and seventy-six questionnaires, out of 1,532 mailed, were returned unanswered for one of the following reasons: incorrect/undeliverable address, not a forest products manufacturer, or opted out by returning an unanswered survey. Two-hundred and eighty-eight questionnaires, 21.2 percent (adjusted response rate), were returned and comprised the data for this study. Of these, 155 were returned after the first mailing, including those that answered using the Internet, and 133 were returned after the second mailing. Nine respondents chose to access the Internet for completing the questionnaire (Table 3).
Firm demographics
One hundred and twenty-eight questionnaires (44%) were answered by firms identified as 0 to 9 employees, 25 were answered by firms with 10 to 19 employees (9%), 76 were answered by firms with 20 to 100 employees (26%), and 59 were answered by firms with more than 100 employees (20%) (Table 3). Percent response by firm size was very similar to those reported from the 1995 study. In 1995, 40.9 percent were answered by firms with 0 to 9 employees, 16.8 percent by firms with 10 to 19 employees, 23.7 percent by firms with 20 to 100 employees, and 18.7 percent by firms with the most employees--combining the 101 to 500 employees category with the more than 500 employees category (Hansen and Smith 1997).
Measuring educational need
A high ranking Educational Need in a subject area was calculated as one that was important but not well known. A subject area that was ranked high in importance but also ranked high in subject knowledge would not receive a high Educational Need ranking. For example, Safety Regulations was rated high in Importance (3 out of 36) but also rated high in Current Knowledge (2 out of 36) for all of the reporting firms. Therefore, the Educational Needs rating for Safety Regulations was 14 out of 36. Competitive Positioning ranked # 17 in Importance but because of its low ranking (23 out of 36) in Knowledge, it ranked higher as an Educational Need than Safety Regulations, a rank of 4 vs. a rank of 14 for all of the reporting firms (Tables 1, 2, and 4).
A subsample of first respondents was compared with a subsample of second respondents to assess the issue of non-response bias. Independent sample t-tests showed no significant difference (p-values [greater than or equal to] 0.05) between the first 30 respondents and the last 30 respondents. This provides some evidence that the response data are valid across the entire sample population (Steel and Torrie 1960).
For all of the firms, Quality and Process Control was ranked #1 in Importance but also ranked #1 in Knowledge (Tables 1 and 2). Because of the high Knowledge ranking, Quality and Process Control ranked #11 as an Educational Need (Table 4). This compares to an overall ranking of #7 by all of the reporting firms in 1995. It is interesting that Quality and Process Control ranked #1 for Importance and #2 for Knowledge in the 1995 study (Hansen and Smith 1997). Obviously, this subject area remains very important to the Oregon wood products industry. It is also interesting that Total Quality Management (TQM) ranks higher as an Educational Need then Quality and Process Control, 8 vs. 11 (Table 4). TQM also ranked higher than Quality and Process Control as an Educational Need in 1995, 6 and 7, respectively. A 1988 survey, targeted and designed to assess educational needs of Oregon sawmills, identified Quality Control, Safety, and Manager Training in Communication and People Skills, and Maintenance as the top priorities (Brown and Niemiec 1997) Again, this illustrates that the industry views TQM and Quality and Process Control as very important subjects for their businesses.
FEED