QuickPLS The research behind this exercise
Rasty and colleagues studied 527 Iranian Instagram followers of travel vloggers using consistent PLS, with 5000 bootstrap draws. The model connected behavioural engagement to trust, social ties, shared values and commitment, and considered similarity-related moderators. Explained variation differed across the intermediate outcomes. The present exercise isolates the reflective correction and two pathways; it does not reproduce the paper’s full moderation or prediction analyses.
The walkthrough below is an original teaching study inspired by this methodological question. It does not reproduce the authors’ data or establish that they used QuickPLS. Read the linked paper for its complete design and findings.
The question you will answer
A travel creator notices that frequent interactions do not always become lasting commitment. Does engagement relate to commitment through trust, through a sense of social connection, or through both?
Each block represents a shared underlying factor reflected in three responses. Engagement is an involvement tendency, not a formative sum of likes, comments and purchases. Social connection is a felt relationship, while Trust concerns confidence in the creator. This factor interpretation is the reason to practise consistent PLS.
Do not treat platform activity counts and an attitude scale as the same measurement. A count of comments is an observed behaviour; a three-item tendency to participate is a reflective teaching factor. Changing between them changes the model and may change whether this correction is appropriate.
Step 1. Download the original practice file
The download has 320 complete fictional records: one fictional travel-vlogger follower per row. Its continuous values sit within a 1–7 teaching response range; they are not rounded survey categories. The number of rows makes the exercise convenient to practise. It is not a sample-size recommendation for a real study.
Open the codebook before importing. Check the meanings and direction of the columns, then confirm the row count and numeric values in QuickPLS. Keep case_id only as a row identifier. It is not an indicator, predictor or control. These files have no planned missing entries, so an unexpected blank calls for an import check.
For your own study, distinguish a missing response from a genuine zero and document exclusions before analysis. Reverse a question only when its wording and scoring require it. The practice columns are already oriented as described below.
Open the import
Open a writable project and choose File → Import Data… or Ctrl+I. In Import dataset click Choose data file… and select this lesson’s CSV.
Review the preview
The dialog moves from Choose file to Review & import. Set Input represents to raw data, Header row to 1, separator to comma and Decimal convention to Dot (1.25). Confirm the teaching columns are Continuous. Keep case_id out of the analysis.
Complete the import
Check that the preview separates the columns correctly, then click Import dataset. Wait for Dataset imported and verified. Open the imported data and confirm the expected case count before building the analysis.
On a small screen, swipe the table sideways to see every column.
| Idea or role | Columns | How this exercise uses them |
|---|---|---|
| Engagement | ENG1, ENG2, ENG3 | reflective construct |
| Trust | TRU1, TRU2, TRU3 | reflective construct |
| Social connection | TIE1, TIE2, TIE3 | reflective construct |
| Commitment | COM1, COM2, COM3 | reflective construct |
Step 2. Build the measurement blocks and arrows
A construct is an idea represented by several measured columns. An indicator is one of those columns. Here each three-column block is reflective: the three scores are treated as expressions of the same idea. This is a teaching assumption, not evidence that a newly written questionnaire is valid.
Use the exact assignments in the column table. The planned structural paths are Engagement → Trust; Engagement → Social connection; Trust → Commitment; Social connection → Commitment.
Create the model
With the imported data open, choose File → New Model…. Give the model a distinct name, select this lesson’s file under Dataset version, check Cases and click Create model.
Create the first block
In Indicators, click the first column of the first block. Hold Ctrl while selecting its other two columns. Check 3 selected and click Create construct.
Name it and set measurement
Choose Select and click the construct body. If needed, open View → Show Properties. Under Properties → Model enter the name from the column table and press Enter. Under Parameter → Measurement model choose Reflective.
Add the remaining blocks
Repeat that selection, naming and measurement check for every three-column block. Keep each column in its intended block. To fix an assignment, use Properties → Data Binding → Remove, then select the right column and use Assign to [construct name].
Draw and save the planned paths
Choose Connect. Drag from the body of each source construct to the body of its destination, then release. Follow the arrows listed above and return to Select. Check arrow direction and save the model. An arrow records a proposed relationship; it does not establish cause and effect.
Step 3. Use the correction that matches a factor interpretation
Consistent PLS, or PLSc, adjusts eligible reflective PLS relationships using estimated reliability. Choose it because the indicators are intended to reflect underlying common factors. A common factor is the shared idea that helps explain their association. PLSc is not a button for making an inconvenient path larger.
Choose Consistent PLS
Click Calculate… and search Find a method for Consistent PLS. Confirm the same dataset and indicator assignments you planned.
Record the available settings
Set Weighting scheme to Path weighting, Result data to Standardized and Missing data to Casewise deletion. In Advanced settings set Maximum iterations to 3000 and Stop criterion to 0.0000001 to match this lesson's reference settings. Check the four reflective blocks and four paths before calculating.
Calculate and inspect the correction
Review Readiness and click Start calculation. Open Reliability and validity → PLSc correction reliability, then Discriminant validity → PLSc construct correlations. Read Original and Corrected across the same row.
Read corrected estimates together
Inspect corrected loadings, paths and R-square under Estimates. Use Search result outputs if needed. Review Diagnostics and warnings before interpreting a correction or comparing it with ordinary PLS.
Step 4. Check the corrected structural relationships
Use the reflective model in this lesson with consistent PLS, path weighting, standardised indicators, starting weights +1, a 3,000-iteration limit and stopping criterion 0.0000001. Read the corrected path table rather than borrowing ordinary PLS estimates.
The uncertainty column uses 999 usable common-row corrected fits out of 1,000 requested, with R seed 261107103 and percentile endpoints. This is a separate repeated-fit exercise, not the app’s native bootstrap or BCa command.
About these answers: Corrected paths, R² and weights agreed between QuickPLS calculation code and cSEM. The loading columns did not have established equality. This was not a new run in the installed app.
On a small screen, swipe the table sideways to see every column.
| Corrected path | Point estimate | Common-row 95% percentile interval |
|---|---|---|
| Engagement → Trust | 0.630 | 0.543 to 0.699 |
| Engagement → Social connection | 0.276 | 0.178 to 0.379 |
| Trust → Commitment | 0.423 | 0.335 to 0.511 |
| Social connection → Commitment | 0.289 | 0.193 to 0.379 |
Compared quantities and retained differences
19 point comparisons of paths, R² and weights and 3,996 usable repeated-path comparisons passed tolerance 0.00000001. The reported loading columns differ between these implementations; the raw and corrected values are retained in the download for separate definition review. No loading-equality claim is made.
1 QuickPLS resampled fit was unusable because its rho_A reliability estimate exceeded 1. The calculation rejected it. The failed fit remains in the checking record and is not counted as agreement.
Do not clip an invalid reliability estimate or substitute an ordinary PLS fit for a failed corrected fit. Keep requested, usable and failed counts with the result. This answer checks the named structural quantities, not every PLSc output.
Step 5. Decide whether the corrected result is usable
Read the Trust and Social connection item meanings together. Feeling personally connected to a creator does not imply confidence in the accuracy of their advice. If corrected correlations imply that the factors cannot be distinguished, investigate this overlap rather than interpreting both path labels as independent discoveries.
Check rho_A, the reliability quantity used in this correction, for each block. Inspect whether corrected construct correlations remain mathematically valid. A value outside the correlation range is a problem to investigate, not a number to trim to a convenient limit.
The ordinary and corrected columns describe different estimation targets. Match cases, blocks, paths and preprocessing before comparing them. Paths depend on several correlations together, so do not assume that each corrected path is simply the old path multiplied by the same constant.
Step 6. Use matching PLSc uncertainty
Choose Calculate… → PLSc Consistent Bootstrapping. Match Weighting scheme and Result data to the point calculation. Set Bootstrap samples to 1000 and Confidence level (%) to 95 for practice. Review Hypothesis test and choose a two-sided question; retain Percentile and BCa under Confidence intervals. Seed is in Advanced settings; enter 261107103 and record Maximum iterations and Stop criterion.
Start the run and read the matching corrected estimates with their intervals. Report usable and failed corrected refits. Ordinary PLS intervals do not supply the uncertainty of a PLSc estimate. If the original correction is invalid, more resamples cannot repair it.
Understand what the correction is doing
The calculation attributes part of the weakening in the score relationship to measurement error under a reflective factor model. It does not remove omitted influences such as creator popularity, nor does it correct a response sample that excludes less enthusiastic followers.
If the corrected value becomes impossible, retain the warning and revisit the measurement model. Taking an absolute value, clipping a correlation at 1 or hiding an unstable block would conceal the very information needed to judge the analysis.
Hypothetical example
Illustrative reading example
Suppose the original correlation between two construct scores is 0.48 and both estimated reliabilities are 0.80. Dividing 0.48 by the square root of 0.80 × 0.80 gives a corrected correlation of 0.60. This demonstrates the reliability adjustment; it is not a fitted value from the download or the paper.
Step 7. Read trust and social connection separately
Open the corrected direct estimates for Engagement → Trust and Engagement → Social connection. Then inspect their separate routes to Commitment. A weaker engagement-to-connection coefficient can be substantively meaningful: responding to content may build familiarity with the advice without creating a felt relationship.
Read explained variation for each intermediate factor separately. Strong explanation of Trust does not mean that Social connection is equally well explained. Nor does either R-square show that the model predicts future followers better than a simple baseline. A prediction claim needs a held-out comparison with its own errors.
For the indirect routes, keep the named Trust and Social connection products separate before adding them. A total hides which intermediate experience carries the association. Use the consistent bootstrap outputs for the quantities you report; do not attach an ordinary PLS interval to a corrected product.
What to check when the result is unexpected
If PLSc and ordinary PLS disagree, first confirm the same rows, blocks and arrows. A difference can follow from the factor correction, especially when reliability differs across blocks. Explain the estimators’ targets before judging one number as wrong.
If a corrected bootstrap repeatedly fails, inspect the affected blocks and warnings. A large requested sample count is not evidence of a stable correction. Report usable refits and consider whether the factor distinction and measurement quality are adequate.
If the social-connection route is uncertain, discuss its interval and what engagement actually measures. Do not replace it with Trust merely because that factor produces the stronger path.
Save the original model and the reason for each genuine correction. Keep unsupported relationships in the results you report. A more convincing study explains a weak or uncertain finding and plans better evidence; it does not erase cases, items or paths until every test is favourable.
Step 8. Write a clear Methods and Results account
Methods outline: The indicators were specified as reflective common-factor measures because [reason]. We used Consistent PLS with [weighting, scale and missing-data settings], examined correction reliability and corrected correlations, and obtained matching consistent bootstrap uncertainty.
Results outline: Original and corrected relationships were [comparison and admissibility]. The Engagement → Trust → Commitment pathway was [estimate, interval], while the Social connection pathway was [estimate, interval]. Report differences in explanation across outcomes without claiming untested predictive superiority.
Include the analysed case count, exact file or data version, model diagram, measurement decisions, calculation settings and the uncertainty procedure. Attach the relevant full tables rather than only the strongest row. Explain which decisions were planned and which followed inspection.
For this exercise, call the records simulated teaching data. In an actual paper, replace every bracketed placeholder with the result from the named saved run. Remove any illustrative numbers from this lesson before drafting your findings. Cite the original study for the research context and the relevant methodological references for the analysis.
Step 9. Report corrected relationships with their own uncertainty
Name consistent PLS and justify the common-factor interpretation of the reflective blocks. Report the corrected estimate and an interval obtained from the same corrected method. Ordinary PLS uncertainty cannot be attached to a PLSc point estimate.
This teaching check requested 1,000 shared-row refits and retained 999 usable corrected fits. The displayed percentile intervals describe that reference exercise. Use the actual settings and counts from your own analysis when writing a research paper.
Check your conclusion before you finish
- Give the measurement reason for PLSc rather than choosing it by coefficient size.
- Find Original and Corrected in the same correlation table.
- Keep corrected estimates and consistent bootstrap intervals together.
- Separate trust, social connection, explained variation and predictive accuracy.