**Low-Quality Statements for Textiles Generated during the Focus-Group Discussions**

Zigzag hemstitching which makes it roll Heavy-weight thread used for the hem (thread sits on top of the fabric) Coarser yarns (lesser twist) Low colorfastness Plastic-like hand-feel Embellishment too strong for fabric Stretchy-like edges which are wonky Wavy-like and forms wrinkles that are not supposed to be there Lower thread count per inch (fewer threads used for warp and weft) Woven threads are visible Machine stitched Tacky tape on the edge Uneven and unfinished look (Note: younger consumers may consider this high quality as a more organic look) Fabric residue and dirty-like after-feel Crunchy sound/noise Stiff fabric and fabric that pokes out High shrink ability Pilling or formation of small balls of fluff from looser threads on the fabric surface Many ravelings—threads from a woven or knitted fabric that have frayed or started to unravel Low colorfast (ability to keep the same color without fading or running even if washed, placed in harsh light, exposed to perspiration, or treated with certain chemicals) Vibrancy (muddy coloring is a sign of poor quality, not drab or muted but muddy coloring) Too big (should not be a shawl when wanting a scarf)

Items that were eliminated after consideration as vague, covered by another or combination of statements, or not really relevant to these garments.

Following another review, the authors further removed statements that had similar meaning or interpretation or were opposites (Table 4, items in parentheses), or items that were covered by other combinations of statements or were deemed less relevant for these garments (Table 4). For instance: "Finished edges" was removed because it was an opposite with "Unfinished edges", and "Even stitches" was removed because it had a similar meaning to "Stitching is even and consistent" which was kept. "Scratchy" was removed because it overlapped meaning (similar or opposite) with "smooth", "rough", and "coarse". Finally, terms such as "springy" were eliminated because they were not deemed particularly relevant to the sample of shawls/scarves in the study. The list of 52 statements was reduced to 30 statements (Table 4, \*statements), which were included in the quantitative consumer study. The statements represented both positive and negative characteristics.

The final set of 30 statements was checked against the focus-group statements by the authors. They were chosen based on frequency of use, ability to be understood by most or all focus-group participants, and to ensure that they captured the breadth and essence of the statements determined by focus groups.


**Table 4.** Shortlist of 52 initial quality statements and 30 final (\*) statements for textiles following first and final review by authors.

#### *3.2. Phase 2—Quantitative Study*

Except for Generation Z participants, many of whom were still in college, approximately 80% of the other consumers had completed college, whereas the remaining had completed high-school education. Although this is not representative of the U.S. population, this more highly educated demographic may be more representative of women who have interest in artisan products and may have, on average, more money to spend on clothing [43]. Data are conflicting on the age groups that are most interested in purchasing artisan products. Some authors [42] have suggested that younger consumers are most interested, while others [43] have found that the older the consumer, the more likely they are to purchase artisan textiles. On, average, consumers in this study held average interest in searching out and purchasing organic and natural products, and on average held moderate to higher interest in purchasing artisan clothing (Table 5).


**Table 5.** LS means for consumers' interest in artisan products and willingness to purchase environmentally friendly products (*n* = 196; α = 0.05).

<sup>a</sup> Row means within an attribute with a common superscript did not differ (*p* > 0.05). <sup>1</sup> Nine-point scale; 1 = Extremely Disinterested, 5 = Have no opinion, 9 = Extremely Interested. <sup>2</sup> Nine-point scale; 1 = Strongly Disagree, 5 = Undecided, 9 = Strongly Agree.

#### 3.2.1. Discrimination among Textile Samples

Mean scores showed that how consumers rated the 30 quality statements, or how they scored the textiles based on quality, or purchase interest, significantly differed among the samples (Table 6). For example, whereas Sample 1 was rated as a very-good-quality scarf and participants indicated that they would probably purchase the scarf if the price was reasonable, Sample 10 was rated as a poor-quality scarf which they would probably not purchase even if the price was reasonable. Furthermore, based on the 30 quality statements that were investigated, Sample 1 attained significantly higher ratings than Sample 10 for positive statements such as Overall attention to detail, Stitching is even and consistent, Hems are even, Consistent weave, Evenly dyed, Looks like it would be colorfast, Smooth, Soft, Corners are straight or curve correctly, Drapes appropriately, and Fine weave. On the other hand, Sample 10 received higher ratings for negative statements such as Unfinished edges, Ravels or frays at edges, Coarse, has an Aroma or smell, Edges are wavy or crooked, and Waxy or sticky feel.

**Table 6.** LS means for consumers' ratings for quality, purchase interest, and the 30 quality statements for textiles (α = 0.05).



**Table 6.** *Cont.*

a,b,c,d,e,f,g,h Row means within an attribute/column with no common superscripts differ (*<sup>p</sup>* <sup>≤</sup> 0.05). <sup>1</sup> Five-point scale; 1 = Very Poor quality, 3 = Fair quality, 5 = Excellent quality. <sup>2</sup> Five-point scale; 1 = Definitely would not purchase, 3 = Possibly would purchase, 5 = Definitely would purchase. <sup>3</sup> Seven–point scale; 1 = Strongly Disagree, 4 = Neither Agree/Disagree, 7 = Strongly Agree.

#### 3.2.2. Internal Consistency of Quality Statements

Computation of coefficient alpha for all 30 statements had a raw alpha of 0.53 and further analysis of the data showed that 13 statements showed inconsistency/unreliability among consumers for measurement of quality. Those statements were: Unfinished edges, Ravels or frays at edges, Patterns do not match when printed or dyed, Rough, Coarse, Soft, Holes loose threads raveling in the interior, Embroidery too loose or too tight, Has an aroma/smell, Edges are wavy or crooked, Waxy or sticky feel, Makes a sound, and Dye spots. That suggests that some consumers (a) were more sensitive to certain characteristics (e.g., smell), (b) varied considerably in how important certain characteristics were (e.g., unfinished edges or raveling), or (c) may not have noticed some characteristics based on their handling of the scarves/shawls (e.g., holes or dye spots). That finding suggests that additional analysis with fewer statements would be beneficial. Such analysis is performed in development of questionnaires for measuring such things as science literacy [44].

After removing the 13 inconsistent statements, the remaining 17 statements showed a Cronbach's alpha of coefficient of 0.86, a reasonably high reliability, but further analysis showed that eight additional statements were potentially inconsistent. Those eight statements were: Front and back are similar, Evenly dyed, Looks like it would be colorfast, Smooth, Handmade, No wrinkles except as part of the design, Feels like natural fabric, and Drapes appropriately. Those statements were ones that may differ based on their perception of "Quality" in artisan textiles. For example, in the focus groups conducted for this study, some consumers expected small variations in artisan products such as uneven dyeing or wrinkles, while others thought of those as "flaws". Some consumers noted that it was difficult to determine if a fabric felt like natural fabric, and, in fact, stated that natural fabrics vary so much (cotton is different from wool is different from silk) that such terminology could not be assessed by feeling the fabric.

Thus, a third analysis was conducted with only the nine remaining statements and the Cronbach's alpha coefficient increased slightly to 0.87, a relatively high number, suggesting that the statements were addressing the quality issue. One additional statement, "fine weave", was noted as being inconsistent in this third round so an analysis was conducted after removing that statement. Those results showed an alpha coefficient of 0.87, consistent with the higher number of statements and no statements that were found to be inconsistent in judging quality among the consumers. Alpha remained 0.87 (Table 7).


**Table 7.** Reliability of quality statements remaining in the scale (overall coefficient α = 0.87).

The eight statements that constituted the textile-quality scale included the statements Overall attention to detail, Fabric is durable, Fringe or tassels look like they will last, Stitching is even and consistent, Hems are even, Consistent weave, Fringe or tassels are neat or tidy, and Corners are straight or curve correctly. This approach of assessing the reliability of the textile-quality scale using coefficient alpha has been used previously by multiple authors [45–47].

#### 3.2.3. Textile-Quality Scale (TQI)

Table 8 shows data for the TQI with eight and six statements. It is clear from the data that the TQI differentiated quite well among the various textiles. The TQ with eight statements was slightly more differentiating among these samples but shows the problem of using the eight statements when some of them do not apply. Sample 1, which did not have fringe or tassels, is the highest-scoring product for the TQI based on six statements, but compares less well to other products when using the TQI with eight statements. Low scores for the statement related to fringe, which obviously did not apply to five of the samples, resulted in a reordering of the samples based on TQI for eight and six statements. However, regardless of which index was used, both differentiated low-quality textiles, those with poor attention to detail, poor-quality seams, poor stitching, inconsistent weave, etc.

**Table 8.** The least-squares means for textile-quality indices (TQI) for the 10 textile samples (α = 0.05).


TQI8 <sup>1</sup> = Textile-quality index generated by summing the scores for all the eight terms that were identified by Cronbach's alpha (range 8–40). TQI6 <sup>2</sup> = Textile-quality index generated by summing the scores for six terms (i.e., eight terms minus "fringe or tassels look like they will last" and "fringe or tassels are neat or tidy" quality terms) that were identified by Cronbach's alpha (range 6–30). a,b,c,d,e,f,g Means within a column with one or more matching letters are not significantly (*p* ≤ 0.05) different.

### 3.2.4. Internal Validation of the Textile-Quality Scale

The Pearson correlation coefficient (all participants as a whole sample) for the score for Overall Quality and the TQI based on eight statements was 0.69, a good correlation. Note that all correlations reported were significant, but the size of the correlation is most important. When the Overall Quality scores and TQI were correlated for each individual participant, only 16%, or 31/196 consumers, had R values < 0.55, suggesting that the TQI for each individual consumer was reasonably representative of their overall quality score.

However, half of the ten samples tested in this study did not have any fringes or tassels. Thus, a TQI based on all eight statements included contributions for the "fringe or tassels look like they will last" and "fringe or tassels are neat or tidy" quality statements. This definitely affected scores for products without a fringe or tassels because consumers give a score for the statement that is not meaningful to the product, which can create confusion for the consumers and potential error in the TQI. Using only six quality statements (removing those related to fringe/tassels) to calculate the TQI provided a correlation for the overall data of 0.71, a slight increase from the eight-statement TQI. Of particular note, however, is that only 10%, or 20 of the 196 consumer participants, had R values <0.55 for their individual TQIs. This suggests an overall improvement, and improvement when considering the percentage of consumers for whom the TQI gave a reasonable representation of the Overall Quality score.

The TQI is not needed when consumers can measure quality directly, but that frequently is not the case for artisans, producers, or sellers who must decide whether their textiles are of high quality using more specific criteria. The use of a TQI composed of various statements of the properties of the textiles allows individuals to learn what consumers believe in terms of textile product quality.

#### **4. Discussion**

Durability, and factors that can affect durability such as attention to detail, were found to be important in this study of artisan textiles. Consumers continue to check for comfort and durability when buying clothing [14]. Yotsutsuji [48] reported that key quality aspects of textile-consumer complaints in Japan were centered on the durability of textiles. This stems from the inherent features of the raw fiber (e.g., silk, wool, cotton, synthetic fibers) to be resistant to wear and tear, resistant to structural changes (e.g., shrinkage, and skewing of textile), and appearance changes (e.g., aging, discoloration, fuzzing). For example, silk (e.g., Sample 1) a natural fiber, has been characterized by a higher percentage increase in length before breaking (breaking extension) as compared to other fibers such as cotton (another natural fiber, e.g., in Sample 10) and polyester (synthetic fiber) [49]. However, the type of processing received by the textile can influence the quality of the resulting textile or clothing [50]. Mechanical processing of the finished textile by the artisan, such as weaving (e.g., consistent weave), sewing (e.g., seams are even and consistent), and edge treatments (e.g., corners are straight or curve correctly) clearly indicated quality to the consumers in this study.

The softness of fibers remains a complex subject today in the clothing textile industry. For instance, while textile makers and sellers may attribute higher quality to higher intensities of fabric softness, the reverse is true for consumers, who may not be drawn to lavish softness [14]. Thus, it is no wonder that for the current study, softness was found to be an unreliable value for quality assessment of artisan textiles such as shawls and scarves. For example, Sample 5 was perceived as high in quality but low in softness, probably because of the stiffness of the weave structure. Furthermore, softness is a difficult attribute to describe because it often is a "lack of" a physical characteristic that provides a soft sensory property. That is, softness often, but not always, results from a lack of stiffness, a lack of roughness, a lack of compression resistance, rather than the presence of a certain attribute.

A study by Ezazshahabi et al. [28] indicated that textiles that had consistent weaves and whose stitching was even and consistent were smoother and thus of higher quality (e.g., Sample 1 in this current study) as compared to textiles with inconsistent weaves and uneven stitching which had a rougher hand-feel (e.g., Samples 9 and 10 in the current study).

Today, evenness of yarns remains a key basis for determining textile quality as consumers continue to find textiles whose stitching is even and consistent more appealing [51]. Although unevenness of yarns to a considerable extent is expected when working with natural fibers such as cotton, Barbu et al. [26] noted that the degree of unevenness can be significantly increased depending on processing. The structural aspects of the textile

clearly influence the quality of the clothing as much as the fiber. That is evident in the current study where Samples 1, 4, 5, and 6 all scored highly for durability and quality even though they were made with different yarns (silk, wool, cotton) and weave structures. In contrast samples 8, 9, and 10 also were cotton, silk, or a cotton and polyester blend, but appeared less durable to consumers and had other issues associated with lower-quality textiles. Sukran [52] found that the type of stitching that is applied to a fabric (e.g., seams or hems) has a significant contribution to the fabric's level of bending rigidity. The level of bending rigidity of the textiles in turn affects their appearance, wrinkle resistance, crease resistance, and draping ability [53]. This lower level of bending rigidity due to type of stitching was illustrated by higher consumer ratings for related statements such as Hems are even, Drapes appropriately, No wrinkles except as part of the design, and Corners are straight or curve correctly for Samples such as 1 and 5 in the current study.

Yarns that are even and which are made from fine fibers produce textiles that are soft and drape appropriately (e.g., yarns used to make Sample 1). Additionally, such artisan textiles usually have even and consistent stitching and are likely to have a consistent weave [54]. The level of importance that textile consumers accord quality statements can vary depending on the function and intended purpose of the particular textile clothing. For instance, in hospitals, the key quality characteristics for knitted surgical gowns and face masks could be their durability and breathability, while for airbags used in cars, the key quality statements could include elongation of yarns used to weave the fabric [29]. Ziegenfus [29] noted also that artisan textiles that can endure abrasion without pilling are considered to be durable and are perceived to have higher quality compared to textiles that form lint balls and lose fibers easily due to wear.

Garside [55] identified hydrolysis, light damage, and oxidation as the main sources of fiber and fabric degradation in textiles. Benkirane et al. [34] stated that durability does not just cover the physical endurance of abrasion and colorfastness, but rather also the emotional durability that a consumer associates with the textile. When purchasing textile clothing, consumers usually think about what the color of the textile would be when washed, dry cleaned, ironed, or when exposed to water or light, as all these are known to affect the color of textile [48]. This would thus partly explain the significantly higher respondent rating for quality and purchase interest for Sample 1 which was associated with being durable and more likely to be colorfast as compared to other scarves or shawls such as Sample 10 and Sample 9.

However, Saville [54] stated that the degree of colorfastness of textile clothing is influenced by various aspects including factors related to the process of dyeing such as the type of dye that was used, the specific color that was used and the color strength (percentage of shade) that was applied to the textile. One study that examined the colorfastness of natural dyes made from eastern red cedar sawdust, Kansas black walnut, and Osage orange on wool yarns that were mordanted with potassium aluminum sulfate, and non-mordanted wool yarns showed that pre-mordanted (with potassium aluminum sulfate) wool yarns had higher absorptions for the Kansas black walnut and eastern red cedar sawdust dyes. This indicated that an increase in depth of shade for the yellow color consequently increased colorfastness to light for pre-mordanted wool yarns as compared to wool yarns that were not mordanted [30]. In the same study, colorfastness to laundering increased only for the Osage orange wool yarns that were pre-mordanted. Their findings highlighted the fact that the loss of color of textiles can be ascribed to several aspects (e.g., water, light, rubbing, domestic laundering, and hot pressing). For the current study, it is thus likely that when consumers were asked to rate the samples based on the degree of colorfastness, they might have interpreted the question differently based on their past personal experiences with colorfastness [34]. The ambiguous nature of the colorfastness question had a significant effect on the internal consistency (reliability) of the textile-quality scale, which explains why the statement was dropped.

It would be expected that textiles whose fringes or tassels are neat or tidy and those that look like they will last (Sample 5) would be perceived by consumers to be of higher quality as compared to similar textiles (Sample 9) whose fringes or tassels are neither neat nor tidy and do not look like they will last. However, at the time of this writing, available literature on the quality of textiles based on fringes or tassels is limited. The findings of one study [32] indicated that dyes and materials could impact the quality of tassels of a textile. For example, a combination of ultra-performance liquid chromatography and time-of-flight spectrometry revealed that tassels (all made from silk fabric) that were directly dyed (no mordant added) with cork tree were more resistant to fading or bleeding due to light as compared to similar, yellow-colored tassels that were directly dyed with cheaper-cost turmeric and had a lesser degree of lightfastness.

Liu et al. [56,57] reported that textiles that were scent infused (with lavender) could be more appealing to potential consumers. Although not related to consumer liking, the current findings show that consumer textile scores for quality could be influenced by the smell or aroma the textile possesses. In many cases for the textiles used in this study, any odor present was quite low and was not noticeable to some focus-group participants. This made the textile smell statement inconsistent in assessing the quality of the scarves and is the reason the statement was dropped during Cronbach's alpha computations.

Respondents rated the quality of scarves and shawls that produced a more pronounced sound when rubbed between fingers as lower in quality than ones that made little or no sound. That could be because consumers believe textiles that produce more-intense sounds wear out easily as compared to those that produce no sounds [58].

Yashi [59] stated that, beyond the primary quality values of textiles (e.g., colorfastness, and durability), secondary quality aspects such as aesthetics and characteristics we measured such as Overall attention to detail and Corners are straight or curve correctly, have a significant contribution to consumers' acceptance and willingness to purchase textiles. This would explain why the quality of samples such as Sample 1 and Sample 5 was rated higher as compared to that of samples such as Sample 9 and Sample 10, which scored significantly lower on Overall attention to detail and Corners are straight or curve correctly. According to Webster [13] and Engel [6], the growing fashion trend that favors fast delivery and quantity over quality where certain aspects of the textile production process such as fabric material sourcing, custom dye applications, and quality control (attention to detail) could explain the poor quality of artisan textiles. Based on this, it is likely that the production of samples such as Sample 9 and Sample 10, which were dyed using a batik technique, was done hurriedly, by lesser skilled artisans, and at a lower cost as compared to other scarves or shawls in this study.

Clearly, it is evident that the eight statements identified by Cronbach's alpha were consistent in measuring the quality of artisan textiles. However, in practice, use of the eight-statement quality scale could be limited to assessment of quality of textiles that have fringes or tassels. Conversely, the six-statement quality scale can be used more universally by producers of artisan textiles (whether they have fringes or tassels or not) to develop products in similar categories that have a higher success rate in the US market. It is important to note that the process of development of the six-statement quality scale can be adopted and used to develop related quality scales for other types of textiles and clothing and in different countries and regions.

Alternatively, the implications of this six-statement quality scale could be compared with those of previously developed consumer-based scales for assessing textile products. Unfortunately, at the time of this writing, comparable literature on similar quality scales was non-existent. For example, Renata [60] stated that characteristics such as durability, comfort [61], and function mattered most to consumers when they were asked about quality of clothing and house textiles. However, because no inquiry/analysis was performed into how those characteristics (as a group) assessed quality of textiles, those results cannot be comparable to the current findings.

#### **5. Limitations**

As each scarf/shawl was a single artisan piece, the reuse of pieces throughout the testing may have changed some aspects of the garment. For example, one sample received a small tear that had to be gently repaired in one of the last sessions. However, in most cases the original and ending quality of the textile appeared to be similar. For future consumer studies on textiles, researchers could ensure that two or three identical versions of each textile sample are available so that in case one is damaged during evaluation, it can be swapped out with another. However, artisan products often are unique production items that are not reproduced as with industrial production, which could make sampling difficult if multiple identical items are needed.

In addition, this study was limited to scarves and shawls, which do not represent the range of artisan or larger-scale production textiles that are available. However, the statements found in this study could apply to many other types of products and statements that were specific to a more limited scope of products (i.e., those with fringes) were removed from the final quality index without losing any benefit of the scale. Additionally, certain statements could be added to the index when needed for certain types of products. Furthermore, it would be quite informative to compare the current results with those collected similarly for other textile products. However, at the time of this writing, prior literature on processes used to study quality of artisan textile products from the consumer's point of view was lacking. Therefore, this provided a limited degree for comparison of the current results with related earlier studies. This highlights the need for more research on what quality parameters consumers use when choosing different textile products.

Another limitation of this study was the fact that these data were collected from textile consumers in the Midwest region of the USA. Obviously, this sample of study participants may not be representative of the entire country. However, we believe this study, and in particular the process discussed here, could be a source of insight for future related investigations in other locations.

#### **6. Conclusions**

A list of statements that consumer textile experts found to be important in artisan textiles (in this case scarves/shawls) was determined with 30 of those statements being used to help develop a six-statement textile-quality scale. Those 30 statements, with a focus on the six-statement TQI scale established in this study, can serve as critical guidelines for artisans or process owners in the US textile industry. These specific textile construction and finishing issues should be a focus when training artisans or quality-control technicians of finished artisan textiles. In addition, these statements can be used to explain the importance of every step in the process for people who are responsible for spinning fibers; weaving, knitting, or bonding yarns and threads into textiles; dyeing; sewing; and finishing textiles and fabrics.

Ensuring that textiles, especially artisan textiles, meet all the specifications of the quality scale would improve the product(s) positioning among the targeted textile consumers (higher market success). That can boost textile makers' revenues, especially those of artisans who depend on producing a smaller number of goods and must show consistently high quality. Finally, the process of development of this six-statement quality scale can be applied to other goods in other countries and regions to establish custom quality scales for different types of textiles and clothing.

**Author Contributions:** Conceptualization, D.H.C., E.C.IV; methodology, D.R.S., E.C.V.; software, D.R.S., E.C.IV; validation, D.R.S.; formal analysis, D.R.S.; investigation, D.R.S., E.C.IV, D.H.C., E.C.V.; resources, D.H.C.; data curation, D.R.S., E.C.IV, E.C.V.; writing—original draft preparation, D.R.S.; writing—review and editing, E.C.IV, D.H.C.; visualization, D.R.S.; supervision, E.C.IV, D.H.C.; project administration, E.C.IV; funding acquisition, D.H.C., E.C.IV. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported, in part, by the National Institute of Food and Agriculture, US Department of Agriculture, Hatch, under accession number 1016242.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of Kansas State University with protocol #5930.2.

**Informed Consent Statement:** A notice of informed consent was provided to all participants involved in the study.

**Data Availability Statement:** The data presented in this study are available within the manuscript.

**Acknowledgments:** The authors appreciate the various textile experts and consumers who took time to participate in the various aspects of the study, especially members of the Prairie Star Quilt Guild, Manhattan, KS (USA) who helped make this project a success.

**Conflicts of Interest:** The authors declare no conflict of interest.
