06/10/2026
Question:
I f I I prefer to use There are two possibilities:
Model A — Three higher-order IVs
Training + Communication + Management Support
→ Organizational Factors
Safety Climate + Physical Work Environment
→ Environmental Factors
Self-Efficacy + Work Motivation
→ Individual Factors
:this instead of 7stacking iv. What is the consequences.
Answer:
Yes. I actually prefer Model A for your DBA, provided we can justify the three higher-order constructs properly.
The consequence is not that your study becomes weaker. It means you are making a different modelling decision: the seven specific constructs become dimensions of three broader IVs, rather than seven separate IVs.
The difference
Model A — three higher-order IVs
Training ─────────────┐
Communication ────────┼──► ORGANIZATIONAL FACTORS ──►
Management Support ───┘ │
│
Safety Climate ───────┐ │
Physical Environment ─┼──► ENVIRONMENTAL FACTORS ──►├──► Employee Performance
│
Self-Efficacy ────────┐ │
Work Motivation ──────┼──► INDIVIDUAL FACTORS ─────►
with PSS as the mediator.
So statistically/conceptually you have:
3 IVs + 1 mediator + 1 DV
rather than:
7 IVs + 1 mediator + 1 DV.
---
What are the consequences?
1. Your research model becomes much cleaner
This is probably the biggest advantage.
With seven IVs, your examiner can ask:
> Why training but not knowledge transfer?
Why communication but not leadership?
Why safety climate but not safety training?
Why self-efficacy but not experience?
And suddenly you're defending seven independent relationships.
With Model A, your central research question remains:
> How do organizational, environmental and individual factors influence employee performance?
That is much closer to the original intellectual structure of your DBA.
---
2. Your hypotheses become much simpler
Instead of potentially having:
H1: Training → Performance
H2: Communication → Performance
H3: Management Support → Performance
H4: Safety Climate → Performance
H5: Physical Work Environment → Performance
H6: Self-Efficacy → Performance
H7: Work Motivation → Performance
plus six/seven mediation hypotheses...
you can retain something like:
H1: Organizational factors have a significant relationship with employee performance.
H2: Environmental factors have a significant relationship with employee performance.
H3: Individual factors have a significant relationship with employee performance.
Then:
H4: Organizational factors have a significant relationship with PSS.
H5: Environmental factors have a significant relationship with PSS.
H6: Individual factors have a significant relationship with PSS.
H7: PSS has a significant relationship with employee performance.
And the mediation hypotheses:
H8: PSS mediates the relationship between organizational factors and employee performance.
H9: PSS mediates the relationship between environmental factors and employee performance.
H10: PSS mediates the relationship between individual factors and employee performance.
That is considerably easier to explain at defense.
---
3. But there is a methodological consequence
This is the part we must not gloss over.
If Training, Communication and Management Support are dimensions of Organizational Factors, we are essentially saying:
> These three things collectively represent the broader organizational construct.
The examiner can therefore ask:
> "Why should these three belong together?"
You need evidence and theory to answer that.
Likewise:
Environmental Factors
Safety climate
Physical work environment
Why are those two sufficient to represent the environmental domain?
And:
Individual Factors
Self-efficacy
Work motivation
Why do those two represent the individual domain?
We need to establish this through the literature.
---
4. Your measurement becomes more sophisticated
This is actually where the co-SV's phrase "cleaner model with stronger measurement validity" becomes interesting.
You aren't simply measuring:
> "Organizational Factors"
with three random questions.
You would measure the first-order dimensions:
Training → several questionnaire items
Communication → several questionnaire items
Management Support → several questionnaire items
and these collectively form the higher-order Organizational Factors construct.
So your measurement model becomes:
ORGANIZATIONAL FACTORS
▲ ▲ ▲
│ │ │
Training Communication Management
Support
Similarly:
ENVIRONMENTAL FACTORS
▲ ▲
│ │
Safety Climate Physical
Environment
and:
INDIVIDUAL FACTORS
▲ ▲
│ │
Self-Efficacy Motivation
This is much more academically interesting than simply throwing seven variables into the model.
---
5. There is one danger: don't call them "higher-order" unless you actually model them that way
This is important.
There is a difference between:
> "Organizational Factors consists conceptually of training, communication and management support."
and actually specifying a higher-order construct in PLS-SEM/SEM.
If your eventual analysis treats Organizational Factors as a higher-order latent construct, then we need to decide whether the dimensions are reflective or formative.
And I would not decide that yet.
For example, if you remove training, Organizational Factors could still exist because communication and management support remain. That starts to look like a formative relationship.
But if Organizational Factors causes people to perceive training, communication and management support as manifestations of the same underlying construct, that would suggest a reflective structure.
This is a methodological decision that should be made after we examine the constructs and their established scales, not guessed now.
---
6. There is another advantage for your particular DBA
Your title is:
> Exploring the Antecedents of Mill Worker Performance: A Case Study of the Palm Oil Industry in Carey Island
Your three broad antecedents actually make conceptual sense:
Organizational → Environmental → Individual → Performance
They describe three levels of influence:
Organizational: what the organization provides/does
Environmental: the conditions in which the employee works
Individual: characteristics/resources of the employee
That gives your study a coherent story.
The seven dimensions then provide the specific operational meaning of those three levels.
I think that is stronger than turning the DBA into a study of seven unrelated predictors.
---
My recommendation
I would choose Model A, provisionally.
But I would phrase it carefully:
> Organizational factors, environmental factors, and individual factors are conceptualized as three higher-order constructs, each represented by specific dimensions identified from the literature. Organizational factors comprise training, communication, and management support; environmental factors comprise safety climate and physical work environment; and individual factors comprise self-efficacy and work motivation.
Then we need to do the real academic work underneath it.
For each of the seven dimensions, we establish:
Higher-order IV Dimension Definition Theory Literature Measurement
Organizational Training ? JD-R ? ?
Organizational Communication ? JD-R ? ?
Organizational Management support ? JD-R/SET ? ?
Environmental Safety climate ? JD-R ? ?
Environmental Physical work environment ? JD-R ? ?
Individual Self-efficacy ? JD-R ? ?
Individual Work motivation ? JD-R ? ?
That table is where I think we should go next.
Because once we fill that in, we'll know whether the co-SV's proposed seven dimensions genuinely support the three higher-order IVs—or whether one of them doesn't belong.
And importantly, we don't need to abandon your original model. We're making it more precise. That is exactly the kind of amendment I would prefer after proposal defense: clarify and operationalize what was already there, rather than inventing an entirely new research model.
Norsanina.blogspot.com
