Contents: Introduction • Method • Results • Summary of the patterns in the data • Responses to various categories of 'error' • Discussion of results by occupational group • Results by gender • Appendix: Annotated facsimile of the questionnaire
 

Introduction

In 1981 Maxine Hairston ("Not all errors are created equal", College English, vol. 43, no. 8, Dec. 1981) asked herself the question 'what should our priorities be as English teachers?' Hairston developed a survey in which readers were given an opportunity to rate usages considered incorrect according to their severity. Almost 20 years later, Johanna Rubba and her class English 390: Modern English Grammar at Cal Poly State University in San Luis Obispo, California, undertook an experiment similar to Hairston’s. First, a list of sentences was prepared which exhibited various deviations from written Standard English. Survey-takers rated their level of discomfort upon reading such writing.

This replication of Hairston's study had three purposes:

1 - to compare current results with the results of Hairston's study, to see if particular judgments have changed over time;

2 - to compare judgments across three occupational groups that have influence over the course of a student's career at very different life stages: schoolteachers, college teachers, and individuals in private industry who might be hiring or evaluating others;

3 - to test certain predictions about particular categories of usages considered erroneous.

Method

Hairston constructed a sample of 65 sentences containing usages considered to be incorrect and then asked 101 people to rate their level of discomfort upon reading them. The respondents were given the choices [error] bothers me a lot, a little, or not at all. Of the 101 pollees men outnumbered women 4 to 1; respondents held positions of influence over other people’s professional lives; and acquaintances of Hairston made up 85% of the group. Eighty five percent of the 101 surveys were returned to Hairston.

Rubba constructed a list of 51 sentences, similar in form to Hairston's. Of the 51 sentences, 45 contained usages considered to be incorrect and 6 contained no error at all. Items were developed based in part on Hairston's survey and also on usages considered deviant that Rubba frequently observes in student writing or in published documents (textbooks, newspapers). The test sentences fell into seven categories: 6 containing no error, but containing items about which many English speakers are currently unsure, such as the who/whom distinction and pronoun case in conjuncts like 'the director and I' vs. 'the director and me'; 7 containing a single misspelled word; 5 displaying dialectal differences from Standard English, such as double negation and alternate verb forms ('seen' for 'saw'); 17 containing usages that are currently undergoing change in Standard English (such as use of nouns like 'impact' and 'access' as verbs, or the who/whom distinction); 1 containing usage considered informal rather than formal; 14 displaying a single punctuation fault; and 1 which failed to capitalize the adjective 'French'. A facsimile of the questionnaire is included in the appendix of this report, annotated with the 'error' or mock error targeted in each item. A clear risk of such a survey is that respondents may be responding to something other than the targeted usage in any given test item. There is no way to discover whether or not this took place.

Rubba’s Modern English Grammar class administered the survey to a diverse group of business people, high school teachers, and college instructors. Each of 25 students was assigned the elicitation of up to 10 responses to the survey. Students knew some of their chosen respondents directly, others indirectly and some not at all. Respondents were chosen for their ability to affect other people’s lives either academically or professionally. When taking the survey, respondents were instructed: "for each item, respond by marking only one of these options: a) I see no error in this sentence; b) The error in this sentence bothers me a lot; c) The error in this sentence bothers me somewhat; and d) I see an error, but it does not bother me at all". The people who chose to take part in this experiment were given one week to respond to the fifty-one questions. The students were given two weeks to collect the completed surveys and submit them for statistical analysis. Anonymity was maintained by separating identifying information from the response forms before the data was input and analyzed computationally.

The survey response forms were computer-readable. The data was read into files and stored and analyzed using Microsoft Excel to summarize data and compile correlations between responses and age, gender, and occupation of the respondents. Summary results were also calculated, as well as responses by category 'no error', 'dialect difference', 'punctuation', etc.

Results

208 surveys were returned and processed, 83% of the maximum goal of 250. For full numerical details, please visit the following website:

http://cla.calpoly.edu/~jrubba/390.RESULTS.html

Additional summaries and bar graphs of the results can be viewed at:
http://www.csc.calpoly.edu/~sstaffor/engl390/index.html

Demographic information: The subject pool was almost evenly divided between males (49%) and females (51%). 31% were between 18 and 25; 49% between 35 and 55; 20% were over 55. Therefore we were surveying a relatively young population, which could be of significance, considering that sensitivity to language changes and education about things like grammar and mechanics vary across generations. 7 individuals did not report their occupation. 41% of those reporting were not educators, but employed in the private sector or by the government; 59% were educators. 18% of the total pool reporting an occupation were middle-school teachers, 13% high-school teachers, and 28% were college teachers. 1/4 of the college teachers were English instructors, including teachers of writing. Here is a full list of the occupations reported by the survey respondents:
 

Schoolteachers		College Teachers		Private industry/ government
Elementary	25	Art	2	Air Quality Inspector	1
Middle/ Jnr. High	11	Biology	2	Airlines Manager	1
High School	24	Chemistry	2	Aquatics Coord.	1
Principal	1	Computer Science	2	Art Gallery Director	1
Dean of Students	1	Education	7	Assistant- Pathology	2
Total	62	English	14	Attendant	1
		French	1	Attorney	1
		Geography	1	Banker	2
		History	1	Broker- Real Estate	1
		Kinesiology	1	Chemist	1
		Math	8	Child Care Administrator	1
		Math- TA	3	Chiropractor	1
		Music	1	Clerk- Manufacturing	1
		Philosophy	3	Collegiate Sport Coord.	1
		Physics	3	Controller	1
		Social Science	2	Coroner	1
		Special Education	2	Dental Administrator	1
		Dept. Chair Ed.	1	Detective	3
		Writing Skills	1	Dir. of Operations Aviation	1
		Total	57	Dir.-Chamber of Commerce	1
				Director of Parks and Rec.	1
				Executive Assistants	6
				Family Specialist	1
				Healthcare Specialist	1
				HR Assistant	1
				Human Services Coord.	1
				Instructor- PG&E	1
				Manager- Accounts	1
				Manager- Bank	2
				Manager- Car Rental	2
				Manager- Country Club	1
				Manager- Credit	1
				Manager- Database	1
				Manager Engineering	1
				Manager- Hospital	1
				Manager- Hotel	1
				Manager- Movie Theater	1
				Manager- Operations	1
				Manager- Paint Store	1
				Manager- Restaurant	2
				Manager- Retail	3
				Mechanical Engineer	1
				Owners	5
				Partner	1
				Pilot	1
				Police Officer	3
				Program Director	2
				Project Engineer	1
				Recreation Coord.	1
				Resource Specialist	2
				Retailer	1
				Sales manager	1
				Software Developer	1
				Supervisor- Agriculture	1
				Supervisor- Computer Lab	1
				Supervisor- Front Line	1
				Supervisor- Lab	1
				Supervisor- Recreation	2
				Systems Scientist	1
				Total	82

Summary of the patterns in the data

Responses to various categories of 'error':

Comparative 'scores' among categories of test items: The first number reflects the percentage who marked either b or c (bothered to some degree), the second number reflects the percentage who marked either a or c (sees no error or is not bothered at all by error).

• Dialect difference 89-11
• Punctuation 68-32
• Misspelling 67-33
• Capitalization 60-40
• Variation in the standard dialect 47-53
• No error 40-60
• Informal style 30-70

No-error Category: 50% correctly saw no error; 50% thought the items contained an error. Of those 50% seeing an error, the majority were not bothered very much by the error they thought was present.

Misspellings: Most (78%) noticed the spelling error, and the majority were bothered to some degree by the error (68%). Those not seeing the error or not objecting at all to the error were  a third of the subject pool (33%)

Dialect differences: 94% detected a dialect difference; 90% objected to some degree to this difference. At 78% 'bothers me a lot', this category attracted the largest number of maximum negative responses. A very small portion of the subject pool either did not detect or were not bothered at all by the dialect variation.

Changes underway/variation within the standard dialect: This category has the broadest diversity of judgments. Though 68 detected something that they viewed as an error, 'botheration' levels were about equal for the 'a lot' and 'somewhat' categories. Just over a third of the subject pool did not detect anything they would characterize as an error. A slightly larger number of subjects than for other categories saw an error but were not bothered at all.

Informal style: One item is too little to judge for this category. However, most of the respondents (59%) saw no error in this item. Those who detected an error were not much bothered by it — only 13% were 'bothered a lot'.

Punctuation faults: 80% of the subject pool detected an error, and, as with spelling, most of the subjects (68%) were bothered by the error to some degree. Nearly a quarter of the respondents did not notice the punctuation fault.

Failure to capitalize: Again, too tiny a sample to generalize from; but it patterns similarly to spelling and punctuation. However, the number of respondents bothered only a little or not at all is close to half: 49%.

Discussion of results by occupational group

Overall, the college teachers and business people were the strictest — these two groups had figures of 60% or more when combining the (b) and (c) ratings. The other two groups — middle- and high-school teachers — had combined (b)-(c) ratings between 50% and 60%. The middle school teachers were the most tolerant, including noticing fewer errors than other groups (their (a) rating is the highest of all occupations). Business people were next-least-likely to notice an error, with nearly a third of respondents marking (a) on average. Only a quarter of high-school and college teachers were likely to find no error in any given item. The college teachers were the strictest group, with the highest number of 'bothers me a lot' ratings (though this was not quite half the pool of college teachers).

The college teachers are most accurate in the no-error category, although only on a 60-40 ratio. (That is to say, 40% saw an error where there was none.) Middle-school teachers were next most accurate, followed by business people and then high-school teachers. Note that, where an error was detected, the tendency was towards not being terribly bothered.

Teachers, especially high-school and college teachers, are least tolerant of misspellings. People employed outside of teaching are only slightly more tolerant. The group most tolerant of misspellings is the middle-school teachers.

All groups come down hard on nonstandard-dialect forms in writing, with college teachers leading the pack. The middle-school teachers are the most tolerant, but only relative to the other occupational groups. They are still quite unforgiving of written nonstandard-dialect forms.

College teachers are most likely to see and object to a standard-dialect usage that is undergoing change. The middle-school teachers and business people are somewhat less likely to spot one of these usages.  However, when those in these two groups do consider a usage erroneous, they tend to be bothered by it. High-school teachers are more forgiving, with high rates of tolerance (answers (c) and (d)) even when they consider the usage erroneous.

High-school and college teachers are both more aware of and less forgiving of punctuation errors, but by relatively slight margins. College teachers come out strictest in this category. Once again, the middle-school teachers are the most forgiving, with ratings almost evenly distributed across (a)-(b)-(c). In every occupational group, more than half the subject pool both notices and objects to some degree to punctuation faults.

Within this tiny category, we find all but the college teachers to be very forgiving of informal style in writing. If more items had been included, the ratings might be different.
 

Results by gender

M/F are about equal in the 'I see no error here' category. Women are very slightly more bothered in the (b) category overall. In most cases, men are more likely to vote 'bothers me somewhat' although the variation is slight and there are a few categories in which women respond with higher frequency with this answer than men. With one exception (dialect differences), men are more likely to see an error but not be bothered by it at all. Overall, the gender differences are slight, but where they do show up, in the majority of cases women are stricter than men.

Appendix

Facsimile of the questionnaire

PLEASE DO NOT COMPLETE THIS SURVEY UNLESS YOU ARE 18 YEARS OF AGE OR OLDER. Please read the accompanying ‘Letter of introduction/informed consent’ BEFORE responding to this questionnaire. You may skip any items you prefer not to answer. All of your replies will be kept completely confidential. Answer all questions by filling in the appropriate letter bar corresponding to the number of the item on the accompanying computer-readable response form.

Please indicate your:

1) Gender: (a) Male (b) Female 2) Age: (a) 18-35 (b) 35-55 (c) over 55

3) Occupation: (a) Private industry/government agency (b) Middle-school teacher (c) High-school teacher (d) College teacher

Many of the items below contain usages of English grammar, vocabulary, spelling, punctuation, or capitalization that would be considered mistakes by English teachers. Not all items contain errors, however; therefore, if you see no error in n item, it is perfectly all right to indicate this Read each sentence at normal speed. Please respond to the sentence as you would if it came across your desk in a letter, report, or other formal document coming from someone you would expect to use English at the same level you do.

For each item, respond by marking only one of these options. If you change a response, please erase your original response very thoroughly. You must use a lead pencil to complete this survey.
 

Response options:
(a) I see no error in this sentence.
(b) The error in this sentence bothers me a lot.
(c) The error in this sentence bothers me somewhat.
(d) I see an error in this sentence, but it does not bother me at all.

4) No stockbroker can guarantee that you will not loose your initial investment.
'Loose' is a misspelling. The correct spelling is 'lose'.

5) The new restaurant in the square serves excellent french dishes.
The word 'french' should be capitalized, i.e., 'French'.

6) Our approach to school reform is different than other programs.
After 'different', usage within the standard dialect varies between 'than', 'from', and 'to'.

7) After the injury, the dog was unable to recognize it’s owner.
'It's' should have no apostrophe: 'its' is the possessive form of 'it'.

8) Students who miss the deadline for adding classes must pay a fee to change their registration.
There is no error in this sentence. The targeted usage was 'their', which may trigger number agreement concerns; in this case, however, 'their' correctly agrees with plural 'students'.

9) A pulmonary embolism is when a blod clot develops in the lung.
This item contained a typographical error misspelling 'blood' as 'blood'. It was originally intended as a style deviation targeting the 'is when', which many authorities find overly informal or outright incorrect in definitions. Preferred would be 'A pulmonary embolism is a blood clot which develops in the lung'.

10) The new law would force all instructors to join the teachers union.
Should there be an apostrophe in 'teachers', i.e. teachers' (plural possessive) or teacher's (singular generic possessive)? I have observed this use of a plural noun as noun modifier on many occasions in newspapers and magazines, but my intuition tells me that the word is a plural possessive and thus requires s'.

11) I seen more than one accident at that intersection.
This sentence displays dialect variation in the past-tense form of the verb: seen where standard English has saw. Many dialects of English (usually associated with the working class or residents of rural areas) switch the past tense form with the past participle or vice versa. In general, principal parts of verbs vary widely across the thousands of dialects of English.

12) We will contact you when a decision has been made, please do not contact us before then.
This item displays a 'comma splice'. The comma between 'made' and 'please' should be a period (ful stop) or a semicolon.

13) A short list of candidates for the job will be compiled from those who we interview.
According to more-conservative varieties of English, 'who' in this sentence should be 'whom', since, in the relative clause 'who we interview', 'who' has the role of direct object, and 'whom' is the direct-object form of this pronoun. However, the who/whom distinction is breathing its last gasps, as English continues the trend, begun over a thousand years ago, of eliminating case-marking.

14) We thought it best to decline the lavish gifts sent to the director and me.
There is no error in this sentence. The targeted usage was 'the director and me' (compare ' sent to me'). Case form in pronouns that are part of a conjunct with 'and' or 'or' is an uncertain matter among educated speakers of standard English. Hypercorrection to avoid 'X and me' in nominative positions leads to use of 'X and I' in all positions, even non-nominative ones as we see here. Respondents who feel uneasy whenever they hear or see 'X and me' may have felt that it should be 'X and I'. As with 'who/whom', the current population is rapidly losing its intuitions for when the objective case forms are called for in certain constructions.

15) High gas prices do not seem to effect consumer demand one way or the other.
Misspelling: 'effect' should be 'affect'.

16) All of the passengers, who were injured in the bus mishap, were transported to the emergency room. The rest left the scene after being interviewed by the police.
Incorrect use of commas around a nonrestrictive modifier: the commas around 'who were injured in the bus mishap' lead the reader to infer that all the passengers were injured, but the second sentence contradicts this inference, making it clear that the relative clause is a restrictive modifier separating those injured from those not injured.

17) Although I followed the manual’s instructions to the letter, I was not able to access my voice mail.
'Access' started life as a noun; in very recent times, it has begun to be used as a verb.

18) In addition to my extracurricular activities in college, I was active in numerous community groups(please see the enclosed resume).
Punctuation fault: there should be a character space between 'groups' and the opening of the parentheses.

19) Apple has once again posted larger than expected profits.
Punctuation fault:'larger-than-expected' should be hyphenated, since it is a phrasal pre-noun modifier.

20) The tornado left a broomhandle firmly imbedded in the trunk of a huge oak.
Misspelling: 'imbedded' should be 'embedded'.

21) Our new fax machine is missing two buttons in its keypad.
There is no error in this item. We do recognize, however, that some may feel that 'from' is a more perspicuous preposition choice than 'in', or that some respondents may have felt that 'keypad' should be written as two words or hyphenated.

22) These second graders’ grasp of spelling rules is very impressive.
No error. There are two targeted 'confusing usages': the correct plural possessive form of graders' ('these' makes clear that 'graders' is plural) and the subject-verb agreement. Singular 'is' is correctly in agreement with 'grasp', but 'rules' comes closest to the verb. Standard English agreement for subjects that are long phrases is shifting from agreement with the 'simple subject' to agreement with the nearest noun to the verb, most likely as a result of short-term memory limitations on language production.

23) The doctor hisself guaranteed that the surgery would solve my problem.
This item displays a different form of the reflexive pronoun: 'hisself' is typical of numerous nonstandard dialects, while standard English retains 'himself'. Note that 'hisself' preserves the logic of the system of reflexive pronouns (possessive pronoun followed by 'self', as in 'myself, yourselves'). As nonstandard dialects often regularize irregular paradigms, this is not a surprising form.

24) We believe that certain measures-including sensitivity training for males and assertiveness training for females-could reduce the incidence of date rape.
Punctuation fault: hyphens should be dashes.

25) The police have not yet determined who was responsable for the vandalism.
Misspelling: 'responsible' is correct.

26) We have made several important business contacts in the Peoples Republic of China.
Punctuation fault: People's is plural possessive.

27) Everyone had paid for their concert ticket in advance.
Singular generic 'their', as a possessive-adjective form replacing 'everyone', is prohibited in traditional grammar. This prohibition was introduced by prescriptive grammarians of the 18^th century. English written usage before that time, including major authors of the Western Canon, contains many examples of 'they' used as a singular generic. The usage has probably never ceased in spoken English, and is making a comeback with the recogniton of gender bias in generic 'his' and the stylistic clumsiness of 'his or her'.

28) Those joining the evening hike should bring the following items; a flashlight, a quart of drinking water, a warm jacket, and a hat.
Punctuation fault: semicolon should be a colon (:).

29) After the meeting, the CEO invited my partner and I to his home for cocktails.
'My partner and I' should be 'my partner and me'. See #14 for details.

30) The Anglo-Saxon’s were firmly established in England by 600 A.D., about 200 years after they had begun invading the island of Britain.
Punctuation fault: No apostrophe needed in the plural Anglo-Saxons.

31) The second sentence of this document infers that you will not assume liability for legal fees.
As with many verbs indicated reciprocal actions (teach/learn, lend/borrow), 'infer' is taking on the meaning of its counterpart, 'imply'. This is very common usage among educated people under 30.

32) When Mr. Dingle come home last night, there was a strong smell of alcohol on his breath.
This sentence displays dialect variation in the past-tense form of the verb: come where standard English has came. Many dialects of English (usually associated with the working class or residents of rural areas) switch the past tense form with the past participle or vice versa. In general, principal parts of verbs vary widely across the thousands of dialects of English.

33) School violence has yet to impact this rural community.
'Impact' started life as a noun; in recent times, it has begun to be used as a verb.

34) The police will arrest whomever is caught entering the condemned structure.
'Whomever' should be 'whoever'; it is the subject of the subordinate clause 'whoever is caught entering the condemned structure'. See #13 for comments on 'who/whom'.

35) Our executive vice-president will soon inform we candidates of the individuals chosen for the overseas task force.
'We' should be 'us', as direct object of 'inform'. There is confusion when a noun complement immediately follows a pronoun.

36) I am trying to contact the person who’s car sideswiped mine on Grand Avenue last February 19th.
Misspelling and punctuation fault: who's should be whose.

37) Every one of my grandmother’s valuable antiques were damaged in the fire.
Subject-verb agreement: 'were' should agree with 'one', not 'antiques'. See #22 for explanation.

38) I have remained loyal to this company. Although the company has not always treated me well.
Punctuation fault: The period after 'company' should be a comma. The string beginning with 'although' is not an independent clause.

39) There’s numerous flaws in your plan to improve assembly-line efficiency.
Subject-verb agreement: 'flaws' dictates that 'there is' should be 'there are' Spoken English is moving rapidly towards having an invariant existential predicate 'there's'.

40) There must be at least one person trained in first aid between the five of us.
'Between' should be 'among'. 'Among' is falling out of use.

41) The new curriculum contains units on phonics and grammer.
Misspelling: 'grammer' should be 'grammar'.

42) This group’s task is to choose the employees whom the President will greet during his visit to the company.
No error; target was 'whom', which is direct object of 'greet'. See #13 for explanation.

43) If our branch was the sales leader, we would qualify for generous yearly bonuses.
Conservative varieties of English use 'were' where this item uses 'was'. 'Were' is the descendant of the ancient subjunctive verb forms that English has been losing for over a thousand years. We are shifting to a system of using the simple past tense in sentences with hypothetical or irrealis meaning.

44) The student group called ‘The Skins’ was originally suspected of setting off cherry bombs in the boys’ bathroom, but it turned out not to be them.
Techinically speaking, 'them' should be 'they', as nominative case is required in this position. But English speakers have been moving away from using nominative pronoun forms in final position, a trend found in other languages (compare French c'est moi, not c'est je, with English 'it is me' vs. 'it is I').

45) I am dissatisfied with your product, but I do not expect no refund.
This item contains a double negative: 'do not expect no'. Double negatives are fine grammar, as long as you are speaking Spanish or French. In English they are found in dialects spoken by disfavored groups, hence they are disfavored. Multiple negation was standard in English until the 17^th-18^th centuries.

46) Before using this device, it should be thoroughly cleaned.
Dangling modifier: who is doing the 'using'? As it stands, the sentence suggests that the device will be using itself, since the modifier 'before using this device' is followed by 'it', which a reader will take to be modified by the phrase. Rewrite: 'This device should be thoroughly cleaned before use'.

47) There is a large number of protesters blocking the main entrance.
No error. 'Number' is singular and is the 'simple subject', hence 'is' is correct.

48) Our new ad campaign promotes Hana Ho as the most unique luxury resort in America’s southwestern states.
Some authorities doggedly maintain that 'unique' cannot be modified with a degree word such as 'very', since it supposedly means 'one of a kind'. But the meaning of the word is changing to be synonymous with something like 'unusual' or 'different'. Such words are compatible with degree words.

49) If your company had informed me of the penalty in a timely manner, I would of paid the charges well before any late fees applied.
Misspelling: 'would of' should be 'would have'. The contracted forms in pronunciation sound identical.

50) Even relatively mild lung infections, can turn into bronchitis or pnuemonia.
Punctuation fault: there is a comma between the full subject 'even relatively mild lung infections' and the auxiliary verb 'can'.

51) The administration is scrapping the program, which is bad for morale.
The comma after 'program' makes the sentence ambigous: is the program bad for morale, or the scrapping of it? If there were no comma, it would clearly be the program which is bad for morale (for those of us who know the comma rules!)

52) The data does not support the claims of the dietary-supplement industry.
'Data' is historically the plural form of singular 'datum'. 'Datum' is very rarely used, especially in laypersons' prose, and 'data' being construed as a singular mass noun like 'sand' or 'gravel'.

53) Patients coming into the office should set themselves down and wait until the receptionist asks for their names.
'Set themselves down' is either colloquial or nonstandard dialect. Standard usage would be 'sit down'.

54) Every year, less students vote in general elections.
A widespread change currently underway is loss of the distinction between 'less' and 'fewer'; whereas 'fewer' used to be used in front of plural count nouns (such as 'students'), and 'less' in front of mass nouns ('fewer bills = less money'), 'fewer' is being phased out altogether and 'less' is taking over This is another change very widespread in individuals under 30.
 

Principal investigator contact information:
Dr. Johanna Rubba
English Dept., Cal Poly State Univ.
San Luis Obispo, CA 93407
Tel. 805-756-2184 or 805-756-2596
E-mail: jrubba@calpoly.edu