What 197 South African workers revealed
More than seven in ten employees in a South African small-business study were classified into a middle group for satisfaction of three basic psychological needs at work. Smaller groups reported consistently higher or lower satisfaction.
The research, published in Frontiers in Psychology on 12 October 2026, examined whether people who actively reshape their jobs also tend to feel more autonomous, capable and connected to colleagues.
North-West University researcher Lynelle Coxen surveyed 197 employees of small and medium enterprises. The findings describe associations at one point in time, not proof that changing a job causes greater motivation.
Three groups, rather than one average
The analysis identified a moderate need-satisfaction group containing 139 employees, or 70.6% of the sample. A lower-satisfaction group contained 35 people, or 17.8%, while 23 people, or 11.7%, belonged to the high-satisfaction group.
These categories were produced statistically from questionnaire responses. They were not diagnoses or fixed personality types, and they should not be applied to every South African worker.
The three needs tended to rise and fall together. The study did not identify a prominent group with high autonomy but consistently low competence, for example.
Why psychological needs matter at work
The study draws on self-determination theory, which proposes that autonomy, competence and relatedness help explain the quality of people’s motivation.
Autonomy means having a sense of choice in how one acts. Competence means feeling able to meet challenges and develop skills. Relatedness means feeling valued and connected to other people.
An employee can be paid for completing a task yet also find it interesting in its own right. That enjoyment is called intrinsic motivation, the study’s principal motivational outcome.
The research asks whether the three needs cluster into recognisable profiles and whether employees in those profiles differ in how intrinsically motivated they feel.
What the survey measured
The 197 respondents worked for South African enterprises employing between 11 and 250 people. Participants were recruited through purposive and snowball sampling rather than a nationally representative random survey.
The sample included 131 women and 66 men. Ages ranged from 18 to 73 years, with an average of 38.1 years. Around 73.6% worked on site, 18.3% in hybrid arrangements and 8.1% online.
Participants completed established questionnaires covering need satisfaction, job crafting and work motivation. The researcher then used confirmatory factor analysis, latent profile analysis and classification-adjusted comparisons.
Because all main measures came from the same survey period, the results cannot determine whether job crafting came before need satisfaction or whether more motivated employees simply reported doing more crafting.
What job crafting means
Job crafting describes employees’ own efforts to adjust their work, rather than waiting for a manager to redesign a role.
The study distinguished three forms. Task crafting changes the scope or organisation of activities. Cognitive crafting changes how someone understands the purpose of their work. Relational crafting changes interactions with colleagues or other people.
These adjustments can be modest. An employee might improve a recurring process, connect a routine duty to a wider goal, or build a more supportive working relationship.
The study measured reported behaviours, not a formal workplace intervention. It therefore cannot establish which specific adjustment would improve an individual employee’s experience.
Task and cognitive crafting stood out
In classification-adjusted analyses, employees reporting more task crafting were more likely to belong to the high rather than moderate need-satisfaction profile.
Cognitive crafting showed a similar pattern. The reported odds ratios were 11.29 for task crafting and 26.40 for cognitive crafting in the high-versus-moderate comparison.
These are model-based odds ratios for questionnaire scores, not percentage increases in motivation. They should not be interpreted as guarantees that a particular action will move an employee into a different group.
The estimates also came from a relatively small sample, so their precision and practical interpretation require caution.
Relationships mattered differently
Relational crafting was associated with membership in the moderate rather than low need-satisfaction group, with an odds ratio of 3.40 in the main model.
Cognitive crafting was also associated with that moderate-versus-low distinction, with an odds ratio of 12.70.
Relational crafting did not significantly distinguish the high and moderate groups in the reported main comparison. This suggests that the three kinds of crafting were not interchangeable statistical predictors.
The differences do not establish that building relationships is less valuable than adjusting tasks. They describe the particular comparisons supported by this dataset.
The largest estimates need care
The original classification-adjusted model produced particularly large odds ratios when comparing the high and low groups, including 71.72 for task crafting and 335.42 for cognitive crafting.
Those figures sound dramatic, but the corresponding confidence intervals were very wide. Small groups and sparse data can produce unstable estimates.
The author ran sensitivity analyses using penalised logistic regression. In those checks, none of the predictors significantly distinguished the high from low group.
The more reliable interpretation focuses on the moderate-versus-low and high-versus-moderate patterns, which were more stable across specifications.
Motivation rose across all three groups
Intrinsic motivation differed significantly among the low, moderate and high need-satisfaction profiles. The overall adjusted comparison was reported as chi-squared 214.09 with two degrees of freedom and p below 0.001.
On standardised motivation factor scores, the estimated means were minus 0.97 for the low group, 0.02 for the moderate group and 1.30 for the high group.
All three pairwise comparisons were statistically significant. Employees with higher satisfaction of autonomy, competence and relatedness also tended to report greater enjoyment of their work.
This is an association between measured experiences. It is not evidence that assigning someone to a high-satisfaction category would automatically raise their motivation.
A strong link with autonomy
The correlation between autonomy satisfaction and intrinsic motivation was 0.85 in the study’s factor-score analysis. Competence was correlated with motivation at 0.69, while relatedness was correlated at 0.47.
These associations were statistically significant, but correlations do not isolate causes. Employees’ job characteristics, managers, personalities and working conditions may contribute to several measures at once.
The result is consistent with the theory that having meaningful choice, feeling effective and experiencing connection all relate to motivation.
It also shows why a single average motivation score may conceal substantial differences in how people experience their working lives.
Why the middle group is important
With 139 of 197 participants in the moderate group, the most common pattern was neither severe dissatisfaction nor exceptionally high fulfilment.
That matters because organisational discussions often contrast highly engaged employees with those who are struggling. This sample contained a large middle group with room for different experiences.
However, the classification is relative to the people and measures in this study. The 70.6% figure must not be presented as a national estimate of South African employee satisfaction.
Future research would need larger, more representative samples to establish how common similar profiles are across sectors and regions.
The context of small businesses
Small and medium enterprises may offer flexibility and close working relationships, but they can also have fewer formal systems for training, progression and role design.
The researcher examined these workplaces because employees may sometimes need to take a more active role in shaping tasks and relationships.
Participants worked across sectors. Finance, real estate and business services formed the largest category at 34.0%, followed by community services and other industries.
The sample therefore captures more than one occupation, but its non-probability recruitment still limits how confidently results can be generalised.
How the statistical groups were checked
Latent profile analysis compares models that divide participants into different numbers of groups based on their response patterns.
The researcher evaluated alternative solutions and selected three profiles based on fit, group size, classification quality and interpretability.
The selected model had an entropy value of 0.87, indicating relatively clear classification. Average assignment probabilities for the low, high and moderate groups were approximately 0.96, 0.93 and 0.94.
These figures support the internal separation of the profiles, but do not demonstrate that exactly three such groups exist in the wider workforce.
The questionnaires were reasonably consistent
The study checked whether the survey items measured the intended constructs. Reliability estimates across the scales ranged from 0.77 to 0.94.
A confirmatory factor analysis supported treating autonomy, competence and relatedness as distinct dimensions rather than collapsing them into a single undifferentiated measure.
The reported three-factor structure fitted the data substantially better than a one-factor alternative.
That measurement work strengthens confidence in the analysis, although self-reported responses can still be affected by mood, interpretation and common-method bias.
What managers might take from this
The findings suggest that conversations about employee motivation should go beyond pay and workload alone. Opportunities for choice, mastery and constructive relationships may be relevant.
Managers could explore whether workers have scope to improve tasks, understand how their work contributes to broader goals and build supportive connections.
Those are possible applications, not interventions validated by this study. The paper did not randomly assign employees to job-crafting programmes or track changes after training.
Businesses should therefore treat the results as reasons to ask better questions, not as a ready-made formula for improving productivity.
The employee’s role has limits
Job crafting focuses on what employees can change themselves. That emphasis should not be used to transfer responsibility for unhealthy working conditions entirely onto workers.
Some constraints require decisions about staffing, supervision, workload or organisational policy rather than individual effort.
Research Today has separately covered how unreasonable work demands were associated with defensive behaviour.
The two studies address different mechanisms, but both reinforce the need to examine the work environment alongside employees’ responses to it.
Connections to changing work
Job crafting also appears in research on AI-supported employment. A previous Research Today report described two contrasting job-crafting pathways among 695 German knowledge workers using AI.
The South African study did not investigate AI tools. Instead, it examined task, cognitive and relational crafting across small-business workplaces.
Comparing these studies helps distinguish the technology employees use from the underlying question of whether they can shape meaningful work.
Neither cross-sectional study can demonstrate that a particular redesign causes lasting improvements in employee wellbeing.
What the study cannot tell us
The research captured 197 people at one point in time, used self-reported questionnaires and relied on purposive and snowball recruitment.
The small high-satisfaction group contained only 23 people, which helps explain why some contrasts had wide confidence intervals.
The findings do not establish whether job crafting causes need satisfaction, whether satisfaction encourages crafting, or whether another factor influences both.
Nor did the study measure objective productivity gains, wage growth or changes in staff retention following an intervention.
The results should be interpreted as informative patterns within the sampled employees, not a definitive account of South African workplaces.
What future research should test
A longitudinal study could measure job crafting before changes in need satisfaction and motivation, helping to establish which experiences precede others.
Randomised workplace trials could evaluate whether structured opportunities for task, cognitive or relational crafting lead to meaningful changes over time.
Larger samples could compare industries, business sizes, provinces and working arrangements while producing more precise group estimates.
Researchers could also combine surveys with manager assessments or objective work outcomes to reduce reliance on a single reporting method.
The bottom line
In a survey of 197 South African SME employees, 70.6% were classified into a moderate psychological need-satisfaction profile, 17.8% into a low profile and 11.7% into a high profile.
Task and cognitive job crafting helped distinguish the high and moderate groups, while cognitive and relational crafting distinguished the moderate and low groups in the main analysis.
Intrinsic motivation was highest in the high-satisfaction group and lowest in the low-satisfaction group, with significant differences between all three.
The study suggests that autonomy, competence, relationships and employees’ ability to shape work deserve attention. It does not prove that any particular workplace change will cause greater motivation.
Source Information
Study title: Basic psychological need satisfaction profiles at work: person-centered links to job crafting and intrinsic motivation.
Author: Lynelle Coxen, North-West University, South Africa.
Journal: Frontiers in Psychology, 12 October 2026.
Design: Cross-sectional survey of 197 South African small and medium enterprise employees; latent profile analysis and classification-adjusted comparisons.
