EVT-2 Expressive vocabulary test (II) Expressive vocabulary test (II) is: Individually administered Norm-referenced Assesses expressive vocab and word retrieval Children and adults 2:6 to 90 years Labeling and synonym Word retrieval evaluated by comparing expressive and receptive vocab skills using standard score differences between EVT-2 and conformed PPVT-4. Each of forms contain three examples followed by 190 items arranged in order of increasing difficulty. Examiner presents a picture from the test easel and reads a stimulus question from the record form. Examinee must respond with one word that provides an acceptable label for the picture, answers a specific question about the picture, or provides a synonym for a word that fits the pictured context Examples are given to the examinee at the beginning to ensure understanding of the task. Key Features development- broad sampling of words, parts of speech, full-colour pictures, representative of races, biased items removed before printing, normed on a nationwide sample of 3540, matched to US census data with regard to sex, race, ethnicity, geographic region, education level (of parents and examines if adult). Administration: start items, basal and ceiling restrict test to areas of ability Examinee not required to read, write, or give lengthy oral responses. Correct responses are listed in report forms Age-based or grade-based comparison. Standard scores allow comparison to other tests with same: i.e. tests of oral language, academic achievement, and cognitive ability Percentiles, normal curve equivalents, stanines, and age and grade equivalents also available Growth scale value Applications: screening, progress monitoring, research by speech-language pathologists, reading specialists, psychologists, who measure expressive vocab. Skills in children through to adults. Screening: possible language difficulty Screening preschoolers: high-frequency words Word retrieval: if were able to identify words in PPVT, may be a problem such as aphasia (loss of ability to produce or comprehend language), or other expressive language difficulties (apraxia: loss of ability to make the movements necessary to create speech, dysarthria: difficulty with articulation of speech sounds). Applications cont’d Reading difficulties: understanding and assessing literacy skills Monitoring growth: GSV Research: longitudinal, clinical ELL: Inappropriate to report a normative score for someone who is not proficient in English. However, the test can be used to evaluate examinee’s knowledge of words and their type (high-frequency, etc). Screening and intervention. Quantitative vs. qualitative analysis Administration of the test Qualifications Rapport Special needs Administration cont’d Order of administration: PPVT then EVT. Why? PPVT is easier, speech not required. Same order as in standardization. Forms: A and B. If you want to compare same items on both, use A then B, or B then A. If you don’t need to compare, use A for both or B for both. Seating arrangement: unobstructed view of examinee, right angles, don’t show answers or report form. Recording info: verify all info, esp. birth date. Reason for administering: assessing word retrieval skills, expressive vocabulary, or screening for language impairment, depending in reason for testing. Why do you need the birthdate? Start item and converting raw to standard scores. 30 days in a month, don’t round days up to next month. Administration 1. Read instructions for first examples on easel, depending on age of examinee (1/2 for 2-6:11 year olds; 2, 3 for 7 and older). If both are correct, go to appropriate start item for their age. 2. If young examinee responds incorrectly to either or both examples, go to first test item. 3. If older examinee has difficulty with 2/3, go back to example one, and begin starting at an earlier start item than their age would suggest. 4. Teach on the start item as well. IE reinforce correct answer after you record it, and if incorrect answer or don’t know, record THEN teach correct answer. “Zipper. This is a zipper.” 5. Show pictures for all questions. For questions that point to a specific body part, if the examinee does not understand, point to your own body part. Pictures also give context for synonyms. Follow the script in the easel and record form by word! Admin. cont’d 6. Basal: five consecutive correct items. Ceiling: five incorrect items in a row. Not including examples. 7. If examinee gets any of the first five wrong, go back til you get to five right in a row. Then jump ahead to where you left off and continue til you get to five wrong in a row. That is your ceiling. 8. If you don’t find five in a row that are right, still move forward to get the ceiling. 9. Administration rules: read the question as it is in the record form. no defining the word, showing the question, spelling of words. Words may be repeated only once. 10. See alternative wording for repeating or prompting: p.11. What do you call this? Tell me another name for what you see. Tell me one word for this. 11. Synonyms: Tell me another word that means the same thing and goes with the picture. Tell me another name for the picture Tell me another way to say..etc. see p 11. Response time: 10 sec. 12. Pointing questions: make sure you point at the right part. Make sure examinee is paying attention when you do this. 13. Same pictures from previous questions: if examinee notices, say yes but this is a different question. And then repeat the question. Some questions have the same picture for two separate questions. Don’t turn the page by mistake. 14. “Pay attention” if distracted. 15. “good job” or “you are doing well” if shy or reluctant 16. if you think he/she misheard, ask them to repeat question back to you. 17. don’t say if it was right or wrong. Just say “that was a good answer”. 18. Remind examinees to look at the picture. 19. Circle/draw line under correct/ incorrect answer on the form. If it is not there, write it in the Other column. 20. DK: don’t know. Prompting P or Figure 2.4 ( p. 13) Only prompt for words in the prompt section of each question Use alternative sentence from p. 11 Another answer that is in the prompt section? Ok prompt again. Don’t overprompt if the answer isn’t coming Remind examinees to say one word only. If the response contains the correct word, and has the same meaning, mark it as right. Others, score according to answer given after prompt. Scoring Correct answers are in the correct box of the record form 1 for correct; 0 for incorrect Record last new response if they change their mind. Don’t suggest or encourage changes if they make a mistake. If you prompt a close answer that is listed in the P box, and they get it right after this, mark it correct. Additional prompts: multiple words, wrong form (part of speech), too vague, related word, too many words, no words, repeat question, wrong meaning for the picture. Scoring rules words are taken from the dictionary. Sometimes two words have one meaning, such as field glasses. multiple word: if contains correct word in same context, mark as right. Don’t penalize for pronunciation or articulation errors (e.g. lisp). Familiarized responses such as “ducky” are okay for young children. shortened words are sometimes okay, such as copter. slang may be okay, such as chopper foreign words are sometimes okay, such as amigo answer must go with the picture. must be specific to the picture. Symbols Brackets [ ]: Multiple answers: remind examinee to restrict to one word answers General categories, e.g. [other animal] Too specific answer e.g. [specific brand], may be wrong Slashes /: alternative words or word endings. May be in correct or incorrect boxes. Parentheses ( ): alternative endings may be correct or incorrect Calculating raw score Same as PPVT: subtract total errors from highest ceiling score. Total errors goes from first error to last. Questions below the basal are counted as correct. Use the highest ceiling possible if there is more than one (rare). Determination of scores Sam: 8 years, 4 months, third grade converting raw score to a standard score: compare to age or grade (Fall/spring) norms depending on examinee’s ability level and results B.1: age B.2: grade fall (July 1 to New Years Eve) B.3: grade spring (New Years to June 30) Form a or B Confidence interval: personal preference and reason for testing. 90 or 95% Same table, B1, 2 or 3. 90% chance that true score falls within the range given (possibility of test error). Percentile, NCE, or stanine: Percentile: B.4 Age equivalent: B.5 Grade equivalent: B.6 Use raw score not standard!!! For age and grade equivalents GSV: B.5 or b.6 Use a or b B.9 for old EVT. extremely low scores: raw score of 0: don’t report standard score. Was test valid for this person? Report behavioral observations rather than standard score Extremely high scores: if they reached the maximum, they may have been able to go further. Note this on observations. Interpreting scores Standard scores: average, above average, below average no bell curve on front; simply a ruler line. Mark an x on the number corresponding to the standard score on the upper section of the number line. Shade in the confidence interval in the lower part. Standard deviations indicate how far the examinee’s standard score was from the mean (average). Significantly different from the norm if 2 SD’s or more. Standard score may remain the same from year to year, but this still means they made progress (as age or grade went up and so did norms). If progress is made, Raw score will increase, so will GSV Percentiles: can’t compare test results from one test to another (ordinal) Stanines: often used as cutoff scores or when broad interpretations of performance are needed (RR- only go from 1 to 9). Age/grade equivalents the age or grade at which the raw score was the average. Should not be used for diagnosis or placement May not reflect individual’s range or depth of ability Cannot compare age grade equivalents from different test. Why? They are not derived scores that are distributed evenly around the mean (i.e. bell curve). Differences between levels are not the same. Don’t misinterpret/overinterpret these scores! GSV as examinee improves over time, GSV will increase. Examine effectiveness of intervention Average can be taken of groups. Score differences between PPVT-4 and EVT-2 USE AGE NORMS!! Need to know their age. go to the back of the report form. Write down the standard score data from both tests on the graph as earlier in the report form (score on top; confidence interval on bottom). Check off which form was used, A or B. subtract scores from each other. Enter difference (positive) in box. check for significance of result in B.7 to see whether this difference was statistically significant or not. Circle the appropriate number. .01 means less than 1% chance that difference happened by chance. Note the smaller the value, the greater the significance of the difference is real and not due to chance. Prevalence and statistical significance Prevalence: how common is this difference in the norm population? B.8: Look up the examinee’s age and difference in the table and look up the percentage in the sample population that had the same difference. This tells you how common/uncommon this is. Circle amount in the Percentage box on report form. “The difference is statistically significant at the .01 level, and only one out of 100 children would have the same difference as his.”= The result is statistically significant (or not) and this is how many people would have the same difference in the sample population. The more significant, the less people will have the same difference. Interpreting score differences If there is no significant difference between scores on EVT and PPVT, and scores are at a similar level, the examinee’s vocab level is consistent receptively and expressively. if there is a significant difference and it is uncommon, look further. if the PPVT is higher, may be a word retrieval problem, rather than expressive vocab. Why? If they knew the answers in the PPVT, they know the answers but are not answering correctly in the EVT. E.g. dysnomia: difficulty naming things. Or other expressive language problem. May want to administer another more comprehensive oral test such as OWLS. Observe child in natural env’t where he speaks. if the EVT is higher, they may prefer expressing their knowledge verbally rather than pointing.