Attitude to Computers

Abstract

Attitude to Computers is a type of attitude — a learned, enduring evaluation — directed at computers and information technology, which MeSH files as a specific case of attitude. It has a tripartite structure: an affective component (anxiety and liking), a cognitive component (beliefs about usefulness and ease of use), and a behavioral component (intended and actual use). It traces two measurement traditions — attitude and anxiety scales, and the capability belief of computer self-efficacy — and the technology acceptance model that recast computer attitudes as predictors of adoption. Debate continues over whether the attitude term adds predictive power once usefulness and ease of use are modeled, and whether results from voluntary workplace software generalize to artificial intelligence. Three demonstrations let the reader score an attitude scale, trace the acceptance path, and watch experience reshape self-efficacy and anxiety.

Keywords: attitude to computers, technology acceptance model, computer anxiety, computer self-efficacy, perceived usefulness

Attitude to computers is where a general theory of evaluation meets a specific, consequential object. The computer is not a neutral tool to the person who must use it: it is approached with a mixture of apprehension and confidence, judged for what it will do and how hard it will be to work, and used or avoided accordingly. Because the object is well-defined and the behavior — use — is measurable, computer attitudes became one of the most heavily studied applied attitude domains of the late twentieth century, and the models built there now govern how the field thinks about every new technology, from the spreadsheet to the chatbot. The subject asks a question that is both ordinary and deep: what disposes a person to embrace or resist a machine, and can that disposition be measured, predicted, and changed?

What Attitude to Computers Is

Attitude to computers is the learned evaluative disposition a person holds toward computers and computing technology — the tendency to respond to them with some degree of favor or disfavor. It inherits its definition from the general construct of attitude, of which it is a specific case: MeSH files it directly beneath Attitude in the behavior and behavior mechanisms hierarchy, marking it as the same kind of psychological object pointed at a particular class of things. What distinguishes it as a topic is the object's clarity and the behavior's measurability. Where many attitudes bear on diffuse targets and loosely coupled actions, a computer attitude bears on a concrete artifact and predicts a concrete, loggable behavior — whether and how much the person uses the system — which is why the domain became a proving ground for attitude theory.

The disposition is not a single feeling but a structured evaluation. A person may be anxious about computers yet believe they are useful, or confident with them yet find a particular system pointless; these are separable, and the separation is the subject of the measurement work below. Crucially, the attitude is acquired. No one is born evaluating computers; the disposition is built from experience, instruction, observation of others, and the direct consequences of use, which is why it shifts over a lifetime and across a population and why interventions can move it. Treating a computer attitude as a fixed temperament misreads it; it is a learned stance that the right experience can revise.

The Structure of a Computer Attitude

The dominant framework imports the tripartite model of attitude: an evaluation is carried by three components — affect, cognition, and behavior — that usually agree but can be measured separately. Applied to computers, the affective component is the feeling the machine evokes, most studied under the name computer anxiety: the apprehension or fear that arises when a person uses or anticipates using a computer. Its measurement was fixed by the Computer Anxiety Rating Scale, which demonstrated that this apprehension is a coherent, quantifiable disposition distinct from general trait anxiety (Heinssen et al., 1987). The cognitive component is the set of beliefs the person holds about computers — above all, how useful and how easy to use they are. The behavioral component is the disposition to act: to approach or avoid, to adopt or reject, operationalized as intended and actual use. Figure 1 shows the three components feeding a single evaluation that then disposes behavior.

Separating the components matters because they can be targeted independently. A person kept from computers by anxiety needs a different intervention from one who is calm but unconvinced of the machine's usefulness, and the general theory of attitudes supplies the reason: the stance that most strongly predicts behavior is the one that integrates all three components and is held with conviction, as the theory of planned behavior formalizes in the path from attitude through intention to action (Ajzen, 1991). Computer attitude research is, in effect, the theory of planned behavior worked out for a single, well-instrumented object.

Measuring Computer Attitudes

Two measurement traditions define the field. The first treats the computer attitude as an object of direct self-report, summing agreement with Likert statements about liking, confidence, and anxiety. The Computer Attitude Scale established the reliability and factor structure of this approach, showing that computer attitudes decompose into stable, replicable subscales — typically anxiety, confidence, and liking — rather than a single undifferentiated feeling (Loyd & Gressard, 1984). The Computer Anxiety Rating Scale did the same for the affective component alone, isolating computer anxiety as a measurable disposition and validating it against behavior (Heinssen et al., 1987). These scales made the construct tractable: once an attitude can be scored, it can be correlated with use, tracked over training, and compared across groups.

The second tradition measures not the evaluation but the capability belief behind it. Computer self-efficacy — the belief in one's own ability to use a computer to accomplish a task — was adapted from Bandura's self-efficacy construct and shown to be a distinct, measurable determinant of use, separable from attitude and from anxiety (Compeau & Higgins, 1995). The distinction is sharp and important: a computer attitude is how a person evaluates the machine, whereas computer self-efficacy is how capable the person feels of operating it. The two are correlated — low self-efficacy breeds anxiety, and competence breeds liking — but they are not the same, and the most predictive models carry both. The first demonstration below lets the reader respond to a short attitude scale, with reverse-scored anxiety items, and watch the subscales and the overall attitude score assemble.

Rate each statement from 1 (strongly disagree) to 5 (strongly agree). Anxiety items are reverse-scored, so every subscale points the same way: higher means a more positive attitude.

result: a mildly positive computer attitude

The scale decomposes a computer attitude into stable subscales rather than a single feeling. Reverse-scoring the anxiety items lets all three combine into one directional total, so a person can be, say, low in anxiety yet only lukewarm in liking — a profile a single number would hide (Loyd & Gressard, 1984).

From Attitude to Use: The Technology Acceptance Model

The most influential development took the computer attitude out of the questionnaire and placed it in a causal chain that predicts adoption. The technology acceptance model proposed that two specific beliefs drive use: perceived usefulness, the degree to which a person believes a system will improve their performance, and perceived ease of use, the degree to which they believe using it will be effortless (Davis, 1989). In the model these beliefs shape the attitude toward using the system, which shapes the intention to use it, which predicts actual use; and ease of use also feeds usefulness, because a tool that is easier to operate is judged more useful. The companion paper tested the model against the theory of reasoned action and showed that this compact belief-attitude-intention chain predicted computer use at least as well as the more general theory, with perceived usefulness carrying most of the weight (Davis et al., 1989).

The model's single most replicated result is the dominance of usefulness over ease of use: across settings, a user's belief that a system is useful predicts intention to use it more strongly than the belief that it is easy. Later extensions kept this core and added its antecedents. The second version of the model, TAM2, specified where perceived usefulness comes from — social influence (subjective norm, image) and cognitive instrumental processes (job relevance, output quality, result demonstrability) — and tested it across longitudinal field studies (Venkatesh & Davis, 2000). The unified theory of acceptance and use of technology, UTAUT, then consolidated eight competing models into four determinants — performance expectancy, effort expectancy, social influence, and facilitating conditions — moderated by age, gender, experience, and voluntariness (Venkatesh et al., 2003), a synthesis its authors later reviewed and extended toward consumer and individual contexts (Venkatesh et al., 2016). The second demonstration lets the reader set perceived usefulness and ease of use and trace the resulting attitude and intention through the acceptance path.

attitude A = 4.50, behavioral intention BI = 5.10 — reduced form BI = 0.70·PU + 0.18·PEOU

Perceived usefulness reaches intention by two routes — through the attitude it shapes and by a direct path — while ease of use reaches it by only one. Collapsing the model gives total effects of 0.70 for usefulness against 0.18 for ease, so a one-point gain in usefulness lifts intention nearly four times as much as the same gain in ease (Davis, 1989).

Individual Differences

Computer attitudes vary systematically across people, and the variation is lawful rather than random. Experience is the most reliable mover: hands-on use raises computer self-efficacy and lowers computer anxiety, so attitudes generally warm with exposure. A study of self-efficacy, anxiety, and attitudes toward the Internet across gender found that reported experience predicted more positive attitudes and that early gender gaps narrowed with comparable exposure, implicating opportunity rather than any fixed disposition (Durndell & Haag, 2002). The pattern is a direct prediction of the self-efficacy account: capability belief is built from mastery experiences, so those with more practice feel more capable and, in turn, less anxious and more positive.

Age is the other heavily studied moderator, and its effect is more subtle than a simple generational divide. The large CREATE program on aging and technology found that older adults' use of technology was predicted not by age as such but by computer anxiety, computer self-efficacy, and fluid cognitive abilities — the same psychological variables that predict use at any age, differently distributed across the lifespan (Czaja et al., 2006). The lesson generalizes: demographic categories like age and gender are proxies for the psychological variables that do the real work, and those variables — anxiety, self-efficacy, prior experience — are themselves the levers. The third demonstration makes the central relationship concrete, letting the reader add experience and watch self-efficacy rise, anxiety fall, and the resulting attitude warm.

at experience 30 — self-efficacy 63, anxiety 47, attitude 58 (a mixed attitude)

Capability belief is built from mastery experiences, so practice raises self-efficacy and, in turn, lowers anxiety and warms the overall attitude. This is why early gaps across age or gender narrow under comparable exposure: the lever is experience, not any fixed disposition (Durndell & Haag, 2002; Czaja et al., 2006).

Worked Example

The technology acceptance model's claim that usefulness matters more than ease of use can be made exact with a simple path calculation. Represent perceived usefulness (PU) and perceived ease of use (PEOU) on a 1-to-7 scale, and take an illustrative path model in which attitude A = 0.5 × PU + 0.3 × PEOU and behavioral intention BI = 0.6 × A + 0.4 × PU. These weights are chosen to illustrate the model's structure, not drawn from a specific dataset; what matters is the structure, in which usefulness reaches intention by two routes (through attitude and directly) while ease of use reaches it by only one.

Substituting the first equation into the second collapses the model to a reduced form: BI = 0.6 × (0.5 × PU + 0.3 × PEOU) + 0.4 × PU = 0.70 × PU + 0.18 × PEOU. The coefficients are the total effects, and they are the point of the exercise: a one-point rise in perceived usefulness lifts intention by 0.70, while a one-point rise in perceived ease of use lifts it by only 0.18 — usefulness is roughly four times as powerful, exactly because it acts through two paths while ease acts through one.

Table 1 works the arithmetic for a user who starts at PU = 6, PEOU = 5 and shows what a two-point gain in each belief buys.

A two-point gain in ease of use raises intention from 5.10 to 5.46, a rise of 0.36. A single-point gain in usefulness raises it from 5.10 to 5.80, a rise of 0.70 — nearly twice the effect for half the change. The arithmetic restates the model's enduring empirical lesson (Davis, 1989): if the goal is adoption, convincing users that a system is worth using beats making it marginally easier, because usefulness reaches intention through both the attitude it shapes and a direct path of its own.

Discussion

Attitude to computers is a textbook case of a general psychological construct sharpened by a well-defined object. Because a computer is concrete and its use is measurable, the domain let attitude theory be tested with a rigor that diffuse attitudes rarely allow, and the payoff was a set of models — the technology acceptance model and its successors — that predict adoption well enough to guide system design and training. The tripartite structure explains why a single number never captured a computer attitude: anxiety, belief, and behavior are separable, and the most predictive accounts keep them apart and then recombine them, exactly as the theory of planned behavior prescribes.

Two tensions remain live. The first is the status of the attitude construct itself within the acceptance models: later versions, including UTAUT, found that once usefulness and ease of use are in the model, the explicit attitude term often adds little predictive power, raising the question of whether attitude is a cause of intention or a summary of the beliefs that cause it. The second is generality. The models were built on the voluntary adoption of workplace software by adults; whether the same weights hold for mandatory systems, for children, or for technologies that provoke qualitatively different feelings is not settled, and the current move to artificial intelligence is the sharpest test of that generality. What is not in doubt is the core finding: computer attitudes are learned, structured, and measurable, and they predict behavior along the path from belief to intention to use.

Current Directions

The acceptance literature has matured into a meta-analytic science. A meta-analytic structural-equation synthesis of the technology acceptance model confirmed, across many studies of teachers adopting digital tools, that perceived usefulness and ease of use predict adoption and that usefulness is the stronger path — now established as a quantitative regularity rather than a single study's result (Scherer et al., 2019). A systematic review of the model in education reached the same structural conclusion while mapping how consistently the core relationships replicate and where they fray (Granic & Marangunic, 2019). The field's center of gravity has moved from asking whether the model holds to measuring how its paths vary across technologies, cultures, and user groups.

The sharpest new front is the migration of the whole framework from computers to artificial intelligence. As machines began to decide and generate rather than merely compute, researchers built and validated instruments for attitudes toward artificial intelligence, finding that these attitudes separate into positive and negative dimensions much as computer attitudes separated into liking and anxiety (Schepman & Rodway, 2020). The continuity is the point: the constructs forged to understand attitudes to computers — a structured evaluation, a capability belief, a usefulness-and-ease calculus feeding intention — are being carried directly into the study of how people evaluate intelligent systems, which is the computer attitude of the present era.

Common Misconceptions

A computer attitude is simply whether someone likes computers.

Liking is only the affective component; a full computer attitude also includes cognitive beliefs about usefulness and ease of use and a behavioral disposition to use or avoid, and these can diverge within one person (Loyd & Gressard, 1984).

Making software easier to use is the surest way to get people to adopt it.

Across studies, perceived usefulness predicts intention to use more strongly than perceived ease of use, because usefulness reaches intention through both the attitude it shapes and a direct path, so convincing users a system is worthwhile usually beats marginal gains in ease (Davis, 1989).

Computer anxiety is a fixed personality trait.

Computer anxiety falls with hands-on experience and rising computer self-efficacy; it is a learned, malleable disposition built from mastery experiences, not a stable temperament (Durndell & Haag, 2002).

Computer attitudes stopped mattering once everyone started using computers.

The same framework now structures attitudes toward artificial intelligence, which separate into positive and negative dimensions just as computer attitudes did; the object changed, the construct did not (Schepman & Rodway, 2020).

Glossary

Affective component.

The feeling side of a computer attitude, most studied as computer anxiety or liking.

Attitude to computers.

The evaluative disposition toward computers and computing technology; a specific case of attitude.

Attitude.

A learned, enduring evaluation that disposes a person to respond favorably or unfavorably toward an object; the parent construct of attitude to computers.

Behavioral component.

The disposition to act toward a computer — to approach or avoid, adopt or reject — usually measured as intended and actual use.

Behavioral intention.

The stated readiness to use a system; in acceptance models the immediate predictor of actual use.

Cognitive component.

The belief side of a computer attitude, chiefly beliefs about a system's usefulness and ease of use.

Computer Anxiety Rating Scale (CARS).

A validated instrument that isolates and measures computer anxiety as a disposition distinct from general trait anxiety.

Computer anxiety.

The apprehension or fear evoked by using or anticipating the use of a computer; the affective core of a negative computer attitude.

Computer Attitude Scale.

A self-report scale establishing that computer attitudes decompose into stable subscales such as anxiety, confidence, and liking.

Computer self-efficacy.

The belief in one's own capability to use a computer to accomplish a task; a determinant of use distinct from attitude and anxiety.

Perceived ease of use.

In the technology acceptance model, the degree to which a person believes using a system will be effortless.

Perceived usefulness.

In the technology acceptance model, the degree to which a person believes a system will improve their performance; the stronger predictor of intention.

Technology acceptance model (TAM).

A model in which perceived usefulness and perceived ease of use shape attitude and intention, which predict use.

Theory of planned behavior.

A general account in which attitudes, subjective norms, and perceived control predict intention and then behavior.

Tripartite model of attitude.

The view that an attitude comprises affective, cognitive, and behavioral components that can be measured separately.

UTAUT.

The unified theory of acceptance and use of technology, consolidating prior models into performance expectancy, effort expectancy, social influence, and facilitating conditions.

Key Researchers

Icek Ajzen

(University of Massachusetts Amherst). Social psychologist whose theory of planned behavior supplies the attitude-intention-behavior chain underlying most measurement and modeling of computer attitudes. Wikipedia · Faculty · Wikidata

Deborah R. Compeau

(Washington State University). Information-systems researcher who adapted self-efficacy theory to computing, developing and validating the computer self-efficacy measure that distinguishes capability belief from attitude. Faculty · Scholar

Sara J. Czaja

(Weill Cornell Medicine). Human-factors and aging researcher who led the CREATE program, showing that older adults' technology use is predicted by computer anxiety, self-efficacy, and fluid abilities rather than by age as such. Faculty

Fred D. Davis

(Texas Tech University). Originator of the technology acceptance model, which recast computer attitudes as perceived usefulness and perceived ease of use predicting intention and use. Faculty · Scholar

Larry D. Rosen

(California State University, Dominguez Hills). Psychologist who built early measurement of technophobia and computer anxiety and later studied the broader psychology of technology use. Faculty · Scholar

Ronny Scherer

(University of Oslo). Educational-measurement researcher whose meta-analytic synthesis of the technology acceptance model quantified how usefulness, ease of use, and attitudes predict adoption across studies. ORCID · Faculty · Scholar

Viswanath Venkatesh

(Virginia Tech). Information-systems researcher who extended the technology acceptance model and unified the acceptance literature into UTAUT, integrating attitude, social influence, and facilitating conditions. Faculty · Scholar

Frequently Asked Questions

What is an attitude to computers?

It is the learned evaluative disposition a person holds toward computers and computing technology, the tendency to respond to them with some degree of favor or disfavor; it is a specific case of the general construct of attitude, applied to a concrete and measurable object (Ajzen, 1991).

What are the components of a computer attitude?

Three: an affective component (computer anxiety and liking), a cognitive component (beliefs about usefulness and ease of use), and a behavioral component (intended and actual use). The components are measured separately and can diverge within one person (Loyd & Gressard, 1984).

What is the technology acceptance model?

It is a model in which two beliefs, perceived usefulness and perceived ease of use, shape a person's attitude toward a system and their intention to use it, which then predicts actual use; it has become the dominant framework for explaining technology adoption (Davis, 1989).

Is perceived usefulness or ease of use more important?

Perceived usefulness is the stronger predictor of intention to use, a result that replicates across many studies and meta-analyses, because usefulness reaches intention both through the attitude it shapes and through a direct path of its own (Scherer et al., 2019).

What is computer anxiety, and is it permanent?

Computer anxiety is the apprehension evoked by using or anticipating the use of a computer; it is a learned, malleable disposition, not a fixed trait, and it falls with hands-on experience and rising self-efficacy (Heinssen et al., 1987).

How is computer self-efficacy different from a computer attitude?

A computer attitude is how a person evaluates the machine, whereas computer self-efficacy is how capable the person feels of operating it; the two are correlated but distinct, and the most predictive models include both (Compeau & Higgins, 1995).

How do age, gender, and experience affect computer attitudes?

They act largely as proxies for psychological variables: experience raises self-efficacy and lowers anxiety, and older adults' use is predicted by anxiety, self-efficacy, and fluid abilities rather than by age itself, with gender gaps narrowing under comparable exposure (Czaja et al., 2006).

Do computer attitudes still matter in the age of artificial intelligence?

Yes; the constructs developed for computer attitudes are being carried directly into the study of attitudes toward artificial intelligence, which separate into positive and negative dimensions much as computer attitudes separated into liking and anxiety (Schepman & Rodway, 2020).

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211. https://doi.org/10.1016/0749-5978(91)90020-T

Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19(2), 189-211. https://doi.org/10.2307/249688

Czaja, S. J., Charness, N., Fisk, A. D., Hertzog, C., Nair, S. N., Rogers, W. A., & Sharit, J. (2006). Factors predicting the use of technology: Findings from the Center for Research and Education on Aging and Technology Enhancement (CREATE). Psychology and Aging, 21(2), 333-352. https://doi.org/10.1037/0882-7974.21.2.333

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. https://doi.org/10.2307/249008

Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982-1003. https://doi.org/10.1287/mnsc.35.8.982

Durndell, A., & Haag, Z. (2002). Computer self efficacy, computer anxiety, attitudes towards the Internet and reported experience with the Internet, by gender, in an East European sample. Computers in Human Behavior, 18(5), 521-535. https://doi.org/10.1016/S0747-5632(02)00006-7

Granic, A., & Marangunic, N. (2019). Technology acceptance model in educational context: A systematic literature review. British Journal of Educational Technology, 50(5), 2572-2593. https://doi.org/10.1111/bjet.12864

Heinssen, R. K., Glass, C. R., & Knight, L. A. (1987). Assessing computer anxiety: Development and validation of the Computer Anxiety Rating Scale. Computers in Human Behavior, 3(1), 49-59. https://doi.org/10.1016/0747-5632(87)90010-0

Loyd, B. H., & Gressard, C. (1984). Reliability and factorial validity of computer attitude scales. Educational and Psychological Measurement, 44(2), 501-505. https://doi.org/10.1177/0013164484442033

Schepman, A., & Rodway, P. (2020). Initial validation of the general attitudes towards Artificial Intelligence Scale. Computers in Human Behavior Reports, 1, 100014. https://doi.org/10.1016/j.chbr.2020.100014

Scherer, R., Siddiq, F., & Tondeur, J. (2019). The technology acceptance model (TAM): A meta-analytic structural equation modeling approach to explaining teachers' adoption of digital technology in education. Computers & Education, 128, 13-35. https://doi.org/10.1016/j.compedu.2018.09.009

Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186-204. https://doi.org/10.1287/mnsc.46.2.186.11926

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478. https://doi.org/10.2307/30036540

Venkatesh, V., Thong, J. Y. L., & Xu, X. (2016). Unified theory of acceptance and use of technology: A synthesis and the road ahead. Journal of the Association for Information Systems, 17(5), 328-376. https://doi.org/10.17705/1jais.00428