QuickPLS Is innovation broader than one kind of change?
A clothing business can innovate in products, production processes or its business model. If those dimensions jointly define sustainability-oriented innovation, one convenient indicator cannot represent the whole concept. The measurement decision comes before asking how innovation relates to circular transition.
Saha and colleagues studied 280 UK textile and clothing firms. Their PLS model used higher-order sustainability-oriented innovation and circular transition constructs and examined contextual moderators. This lesson makes a smaller original model: one higher-order innovation construct predicts one first-order transition outcome. It does not reproduce the paper's complete higher-order or moderation design.
Combine three innovation dimensions and keep the outcome separate
Use 260 fictional clothing businesses. Product redesign, process redesign and business-model redesign jointly form INNOVATION. Each dimension has three invented reflective indicators. TRANSITION is a separate first-order reflective construct with three indicators of perceived progress towards circular operations.
In this teaching definition the innovation dimensions are not interchangeable. A business can redesign a product without changing its production process or commercial model. They therefore form the broader innovation construct. The TRANSIT indicators instead express one narrower perception of transition progress; they are not separate physical and commercial ingredients.
Specify INNOVATION → TRANSITION. This is one higher-order predictor and one ordinary outcome, supported by four original indicator blocks. The workflow creates exactly one higher-order construct in the model. The outcome is deliberately simpler than the published study's higher-order transition construct.
The continuous 1–7 teaching indicators are original, with one row per business. They are not the source questionnaire or objective circularity records. A fitted association would describe the invented perception scores, not demonstrate actual material recovery or revenue from circular products.
On a small screen, swipe the table sideways to see every column.
| Columns | Construct | Measurement role |
|---|---|---|
| PRODUCT1–PRODUCT3 | Product redesign | Reflective dimension of INNOVATION |
| PROCESS1–PROCESS3 | Process redesign | Reflective dimension of INNOVATION |
| BUSINESS1–BUSINESS3 | Business-model redesign | Reflective dimension of INNOVATION |
| TRANSIT1–TRANSIT3 | Perceived circular-transition progress | First-order reflective outcome TRANSITION |
| INNOVATION | Formed from the three redesign dimensions | Higher-order construct; no raw column |
| case_id | Fictional business | Identifier only |
Step 1. Create four ordinary indicator blocks first
In real data, duplicated or misassigned items can make dimensions appear more similar than they are. Resolve content and scoring before choosing a higher-order approach. A high correlation alone cannot tell you whether product, process and business-model redesign should be merged.
Import the data
In a writable project choose File > Import Data… and Choose data file…. Inspect Review & import: raw data, Header row 1, comma separator, Decimal convention Dot (1.25), and Continuous indicators. Choose Import dataset and confirm Dataset imported and verified with 260 cases and twelve measurement columns.
Create the model
Choose File > New Model…, select the imported Dataset version and Create model. In Indicators select each matching three-column block and choose Create construct.
Name and define each block
Use Properties > Model > Name for PRODUCT, PROCESS, BUSINESS and TRANSITION. Assign the TRANSIT columns to TRANSITION. Under Parameter > Measurement model choose Reflective for all four ordinary blocks.
Check assignments
Keep case_id out of the model and check that no indicator is assigned twice. The complete file needs no reverse coding. Save before creating the higher-order predictor.
Step 2. Create INNOVATION, then connect it to the ordinary outcome
A disjoint two-stage analysis first estimates the redesign dimension scores, then uses them to represent INNOVATION. The original indicators remain part of the dimension assessment. The transition indicators retain their first-order reflective role; they are not dimensions to add to INNOVATION.
Select only the innovation dimensions
Select PRODUCT, PROCESS and BUSINESS together. Leave TRANSITION outside the selection. Open Higher-Order Construct from Model or the selection context menu.
Define the higher-order predictor
Name the new construct INNOVATION and choose Dimensions form the HOC. Review the selected dimensions, derived type and recommended approach.
Choose the eligible staged approach
Open Advanced. Under Construction approach choose Disjoint two-stage, then Create. If Initial model path appears for a protected project revision, select the intended path before Create. A fresh model need not show that field. Inspect which dimension scores enter the later stage. Create no second higher-order construct.
Draw the research path
Use Connect to draw INNOVATION → TRANSITION. TRANSITION remains the ordinary reflective construct measured by TRANSIT1, TRANSIT2 and TRANSIT3. Keep generated estimation paths distinct from this research relationship.
Calculate and inspect both levels
Open Calculate… and the supported PLS-SEM Algorithm setup. Review readiness and the higher-order choices, then Start calculation. Open the completed result and read the original measurement blocks before higher-order weights and the structural path.
Step 3. Check the two stages with this practice file
The Disjoint two-stage model first estimates the reflective scores for PRODUCT, PROCESS and BUSINESS without INNOVATION. Those dimensions have no remaining structural arrows in the first stage. Their scores then form INNOVATION in the second stage, while the other constructs keep their original indicators. The stage plan below explains this calculation; create the higher-order construct through the model workflow above.
The broad INNOVATION composite is positively associated with TRANSITION (0.611; reference interval 0.537 to 0.684). This exercise has one higher-order predictor and an ordinary reflective outcome. It does not estimate an interaction, even though the source paper considers a wider design.
The uncertainty column comes from 1,000 usable full-stage reference refits out of 1,000 requested, with R seed 71012. Both stages are rebuilt in every whole-row resample. The endpoints are the 2.5th and 97.5th percentiles.
About these answers: Separate two-stage reference calculation. Component scores and point estimates were compared with QuickPLS calculation code; this was not a new run through the desktop app.
On a small screen, swipe the table sideways to see every column.
| Second-stage path | Point estimate | 95% reference percentile interval |
|---|---|---|
| INNOVATION → TRANSITION | 0.611 | 0.537 to 0.684 |
How the reference answer was calculated
The reference calculated the second stage separately. Its component scores, point paths and R² values were compared with QuickPLS calculation code. This did not run the app's complete higher-order workflow, so it is not an end-to-end check of that workflow. The interval arithmetic was checked against all 1,000 saved full-stage repetitions. Those repetitions were not independently rerun and are not results from the QuickPLS bootstrap command.
Path weighting, +1 starting weights, 3,000 iterations and stopping criterion 0.0000001. For the isolated reflective dimensions, the reference uses the principal-component limit of the own-score iteration, oriented toward the positive item sum.
These reference intervals are not a run of the app’s native bootstrap command. The download retains the sampled row numbers, settings, reference estimates and any failures.
Step 4. Follow the two calculation stages
Use this plan alongside the model diagram. It shows which scores and paths belong to each stage, so an indicator model and a second-stage score model are not confused.
Step 5. Read the dimensions that form the broader concept
A weight describes a dimension’s contribution while the other dimensions remain in the composite. It is not the same quantity as a reflective indicator loading. Keep a theoretically necessary dimension even when its weight needs further investigation.
On a small screen, swipe the table sideways to see every column.
| Dimension | Formative weight | 95% reference percentile interval |
|---|---|---|
| BUSINESS | 0.579 | 0.432 to 0.711 |
| PROCESS | 0.622 | 0.479 to 0.733 |
| PRODUCT | 0.520 | 0.353 to 0.671 |
Step 6. Read dimension quality, formative weights and the outcome separately
In Results use Search result outputs to find Higher-order component relationships, Higher-order structural paths and Higher-order method and run details. The component table separates Loading from Weight. The details table identifies the higher-order construct, components, method and generated measurement; use it to check the specification behind the estimates.
Inspect outer loadings, reliability and average variance extracted for PRODUCT, PROCESS, BUSINESS and TRANSITION. Use HTMT+ to examine whether the redesign dimensions are distinguishable and whether their wording overlaps with the transition outcome.
For formative INNOVATION, read the dimension weights and their collinearity. A relatively large PROCESS weight means a larger conditional contribution under the selected model and scaling. It does not prove that process redesign is the best intervention or that the other dimensions are unnecessary.
Then read INNOVATION → TRANSITION with its uncertainty. A positive broad path does not imply that every innovation dimension has a positive unique direct effect on transition. Those component-specific paths would define a different structural model.
Use HTMT+ for the reflective PRODUCT, PROCESS, BUSINESS and TRANSITION blocks. Assess formative INNOVATION through its dimension weights, collinearity and content. A not-applicable reflective statistic for INNOVATION is consistent with that measurement role.
On a small screen, swipe the table sideways to see every column.
| Output | Research question | Avoid |
|---|---|---|
| Reflective loadings and reliability | Are the dimension and outcome indicators coherent? | Treating high reliability as full content validity |
| HTMT+ | Are the concepts sufficiently distinct? | Merging dimensions solely to improve a table |
| INNOVATION weights | How do dimensions contribute conditionally? | Calling weights causal priorities |
| Stage mapping | Which scores represent INNOVATION? | Treating TRANSITION as a second higher-order construct |
| INNOVATION → TRANSITION | What is the broad fitted association? | Claiming every dimension has this separate effect |
Step 7. Refit the higher-order predictor's stages in the bootstrap
Choose the supported full-model higher-order bootstrap for the saved model. It should refit the required stages using original business rows. Holding previously estimated dimension scores fixed would answer a different uncertainty question.
In PLS-SEM Bootstrapping review the model and selected approach. For the teaching run use Bootstrap samples 5000, Confidence level (%) 95 and Seed 71012 where the controls are available. Start calculation, open the result and record the actual interval method, usable fits and failures.
Read intervals for each formative weight and for the structural path. The path interval can be relatively narrow while one dimension weight remains uncertain. They address different quantities and should not be flattened into one judgement that the entire model passed.
A broad association can coexist with an uncertain dimension contribution
Do not remove BUSINESS and continue calling INNOVATION the same three-dimensional concept. If business-model redesign belongs to its definition, retain it and discuss the uncertainty. A narrower definition is a new model requiring a new explanation.
Hypothetical example
Illustrative reading example
Suppose illustrative innovation weights are PRODUCT 0.45, PROCESS 0.40 and BUSINESS 0.15. Suppose BUSINESS has interval [−0.08, 0.38], while the INNOVATION → TRANSITION coefficient is 0.52 with interval [0.30, 0.69]. The broad association is positive under the hypothetical model, but the separate business-model contribution is imprecise.
The innovation weights are not percentages of a business's activity or value. Their scale depends on the score construction and relationships among dimensions. Likewise, a standardised path of 0.52 does not mean a 52% increase in circular operations.
For the reflective transition outcome, imagine loadings of 0.80, 0.75 and 0.85. Their average squared loading is (0.64 + 0.5625 + 0.7225) / 3, approximately 0.6417. This AVE illustration concerns the outcome's indicators. It is not a reflective reliability assessment for formative INNOVATION.
All weights, coefficients, loadings and intervals in this illustrative example are invented. They separate a broad structural association from uncertain dimension contributions and outcome measurement. They are not fitted CSV values or estimates copied from the paper.
Keep a smaller model honest about the question it answers
Start with the one-higher-order-construct model specified here. Adding another higher-order construct is not part of this creation workflow. If a future study needs a more complex measurement structure, establish a supported analysis plan before writing its app instructions or interpreting its coefficients.
An eligible alternative higher-order approach can be compared if it preserves the intended measurement roles. Record how the stage construction differs and whether weights and the structural relationship change. Keep both results rather than selecting the approach with the largest coefficient.
A reflective higher-order alternative changes the meaning: the broad concept would explain its dimensions. It is not a harmless technical switch for the formative definition. Use it only with a substantive argument.
In real firms, innovation may precede circular transition by years. Cross-sectional perception scores cannot establish that sequence. A positive path therefore remains an association unless the design supports a stronger interpretation.
Locate the problem before simplifying the model
On a small screen, swipe the table sideways to see every column.
| Symptom | Check | Action |
|---|---|---|
| The higher-order selection includes TRANSITION | Selected dimension membership | Select only PRODUCT, PROCESS and BUSINESS. |
| A second higher-order creation is blocked | The one-HOC teaching specification | Keep TRANSITION as the ordinary reflective outcome. |
| BUSINESS has a small weight | Content, collinearity and its interval | Retain the intended definition and explain uncertainty. |
| INNOVATION lacks reflective AVE | Its formative measurement role | Assess formative weights and content; reserve AVE for reflective blocks. |
| The broad path is clear but a weight is not | The distinct quantities being estimated | Report both levels without a blanket pass/fail statement. |
| The result is described as a percentage improvement | Coefficient metric and perception-based outcome | Use score-scale association language rather than an objective improvement claim. |
Step 8. Describe the simplified measurement design accurately
Report the reflective dimension and outcome assessment, the innovation weights and their uncertainty, then the structural coefficient and interval. Explain an uncertain dimension without quietly changing the construct definition. State that this exercise uses one higher-order predictor and does not reproduce the source paper's higher-order transition or contextual moderation.
| Report section | Include |
|---|---|
| Definitions | Why three redesign dimensions form INNOVATION |
| Reflective measurement | Dimension and TRANSITION indicators, loadings and validity |
| Higher-order measurement | One HOC, approach, weights, collinearity and stage inputs |
| Structural relationship | INNOVATION → TRANSITION estimate and interval |
| Scope | First-order perception outcome and original teaching adaptation |
Methods writing outline
We specified INNOVATION as a formative higher-order construct formed by PRODUCT, PROCESS and BUSINESS, each with three reflective teaching indicators. TRANSITION was a separate first-order reflective outcome with three indicators of perceived progress. We used [eligible approach and stage mapping] and estimated INNOVATION → TRANSITION with [full-model bootstrap settings and counts].
Step 9. Explain what this staged model actually tested
Describe the two stages, measurement roles, score scaling and resampling procedure. Use the same model and the same coefficient scale for the estimate and its interval. Report the weights of PRODUCT, PROCESS and BUSINESS separately from the INNOVATION → TRANSITION relationship, and retain the first-order reflective definition of TRANSITION.
Keep the two measurement roles visible
- Name the three dimensions that form INNOVATION.
- Explain why TRANSITION remains first-order and reflective.
- Check all twelve original indicator assignments.
- Distinguish a formative weight from a reflective loading.
- Report an uncertain weight without deleting theory.
- State which parts of the source design this focused lesson does not reproduce.