Mayer et al. (2019) Chapter 8 PDF

Chapter 8 Optimizing Client Progress by Monitoring Behavior Change: Recording, Graphing, and Analyzing Patterns of Change Goals 1. List at least three reasons why graphing behavior is important...

Chapter 8 Optimizing Client Progress by Monitoring Behavior Change: Recording, Graphing, and Analyzing Patterns of Change Goals 1. List at least three reasons why graphing behavior is important. 2. e ne and llu trate the ollo n a. baseline b. datu c. behavioral technician d. o erat onal de n t on e. line graph. y-axis g. ord nate h. x-axis i. abscissa j. vertical phase change lines k. phase label l. bar graph m. u ulat e re ord n. tandard elerat on hart o. change in level p. trend q. variability r. e ual- nter al ra h and d fferent ate the ro tandard elerat on hart s. trendl ne 146  8 147 3. t and d u ho be de beha oral te hn an ht ob er e and re ord behavior. 4. u the ue n ol ed n u n el -re ord n dur n ba el ne and dur n nter ent on. 5. u the rele an e and ro de an e a le o auto at data olle t on. 6. e r be a rule o thu b or deter n n hen reactivity ha ub ded. 7. la n hy t ortant to re ely de ne your ea ure. 8. la n the ortan e o the hedule that you u e n your beha oral- re ord n e on and ho t n uen e the al d ty o your ea ure. 9. u ho lon data hould be olle ted dur n ea h ba el ne and intervention phase. 10. llu trate ho to lot ult le et o d fferent nd o data that are a t to co-vary. 11. fferent ate a on line, bar and cumulative ra h and e y hen to u e each. 12. e r be hen a a change in level and b trend become apparent. 13. e r be hy e a n n the nature and our e o ar ab l ty o ortant. ************* o you onder you are eat n too u h un or other are te ted to on lude that a roble - ood Are you on n ed you et lenty o e er- at beha or ha a led to ro e de te l ht e Maybe a h ld arent and tea her are thor- thou h otent ally n ant rate o han e. hey ou hly er uaded that h beha or totally out o a l to real ze that beha or han e o ten ro re e ontrol the tuat on ho ele. a e u er or radually. et hen re ented th ra h re re- orre t n on lud n that the boy a te too u h entat on o han e n re uen y trend and or t e ur n the eb dur n or n hour o do ar ab l ty o the beha or o ntere t r or to you deter ne the a ura y o hat a tually ha - dur n and a ter a ro ra ha ea ed you and en n Gra h d lay o the baseline—repeated they are better able to dete t any ro e ent measures of the ongoing behavior, designed to rep- that ha e ta en la e. A a beha or analy t you resent its typical pattern of occurrence prior to any u t be erta n to ba e your de on on ob e t e intervention enable you to e the tuat on ore e den e hat the data, the numerical results of realistically. measuring some quantifiable aspect of behavior, tell Beyond n a real t ture o on o n you rather than your ut eel n. beha or ra h al o ro de the ba or de n n en ble and real t interventions, or behavior- change programs, and or on tor n the r effe - The singular form of t ene. hey ob e t ely n or beha or analy t data is datum. about hether beha or ro n or en n or remaining stagnant. Another ortant rea on or u n re e beha oral ea ure ent that t hel to re eal ubtle han e e e ally dur n the n t al ta e o Gra h n e ent al to your ra t e a an an nter ent on. h e e ally aluable be au e a l ed beha or analy t be au e t er e a one not un o only arent tea her ana er and o your o t u e ul d a no t on tor n and 148 analyt tool. or the e rea on on e you ha e nicians M (RBT 1 who implement treatment, ele ted a ro r ate ea ure the ne t te na design and implement observational record- beha or- han e lan are to a tually re ord beha - ing systems, and who are under the supervi- or ra h the re ult and analyze the data under sion of a BCBA. The BCBA supervises the ba el ne and nter ent on ond t on. ere e data collection and graphing schemes for de r be the te you enerally ll ollo hen continual monitoring Sulzer-Azaroff ha re ord n and ra h n and a e you a l ar th ho a 19. Add t onally tea her a e ba te n analyz n ra h data. Be ore ad an ed tudent tea her a de hool be nn n thou h you ll need to re ne your de - y holo t or oun elor arent olle e n t on o the beha or o ntere t and dent y the tudent or arent olunteer are o et e nd dual ho ll ondu t the ob er at on. a ned a data re order. Homes u ually RB but o et e arent IMPLEMENTING or l ent the el e o ten a red or rel - ab l ty ur o e th RB. ollea ue or OBSERVATIONAL SYSTEMS n ant other. Be ore tart n a beha or- han e ro ra you Working in varied locations te hn an ho u t a e the o et e d ult de on about deo-re ord olle t rodu t or other e ho a tually ll re ord the beha or. ll the re er e e den e or later or n o the behavior analyst or the behavior-change agent also beha or o on ern o er t e be re on ble or re ord n data o you ha e the lu ury o taff a ned to that ta ll the l ent n rea n ly any edu at onal and hu an er- re ord h or her o n beha or e a en e a n data olle t on re on b l t e to RB and o et e to Board ert ed A tant Beha or Analy t M B aBA.2 Occasionally, state Behavior-Recording Staff l- er e de art ent ha e bu lt tho e re on - b l t e nto the r hu an- er e areer ladder. dent y n the beha or-re ord n taff de end n hen beha oral te hn an are una a lable art on the ty e o or an zat on th h h you are you u t e lore alternat e. or e a le ar- or n. n lar e enter r e data re ord n ay be ent en or t zen nd dual th de elo en- a e al zed ta a ned to e loyee a an a e t tal d ab l t e ra head Mer ator Bella o the r ob. n our o n e er en e de end n on 19 olle e tudent on ra t u or ntern h the nature o the or an zat on and t re our e ar - a n ent older h ldren M au hl n Mal- ou nd dual n ar ou ett n be de re ear h- aby 19 a eer or e en l ent the el e ee er and the r te hn al taff e ber ha e ul lled belo ay er e a ob er er. n an ele tron th un t on lant to h h one o u on ulted a ret red an- Factories and offices u er or a ety a er a reh red on a art-t e ba to aud t lant taff or eer or er- olunteer. a ety tar et. n another tuat on on tor n and re ord n ual ty ea ure ere n luded n ual ty- Business and service organizations: ual ty- ontrol n e tor ob de r t on Sulzer-Azaroff a uran e taff olunteer u er or l- ar hbar er 199. r n eo le ho no ents. the bu ne an be e e ally bene al be au e Healthcare delivery settings: olunteer or ddle ana e ent er onnel. 1 or n or at on and tra n n re ard n RB ee alla e and Mayer 201. Al o ee the BA B eb te BA B.net Schools and human service agencies: either 2 See the BA B® eb te htt BA B.net or a de r t on o ull or art-t e registered behavioral tech- the re on b l t e o RB and B aBA.  8 149 they under tand the nature o the beha or they are What do we mean by an operational definition in o n to ob er e and re ord. applied behavior analysis? So e ro ra re u re on tor n n any ett n and or around the lo h h ay o e An o er tion l e nition is a clear, precise a art ularly d ult hallen e. hen e ther the description of how the properties of events or contingency managers or the clients themselves items are measured. Operational definitions ay ha e to er e a beha or re order. o e er are essential to the dependable and reliable measurement and implementation of behavioral te hnolo al ad an e ha e ade t o ble to procedures (as in assessing procedural deo-re ord the e on and ub e uently hron - reliability or treatment integrity) and/or the le beha oral data by at h n the deo l. t collection of behavioral data. For a behavior hould be noted thou h that do n o ta e ore to be operational, it must be observable and t e. Al o o et e deo re ord n an re ent measurable. an eth al d le a a n the a e o nad ertently l n oota e o eo le ho ha e not en the r deota e and other auto ated re ord n y te n or ed on ent to be ta ed or e a le the other al o ha e been u ed to ro de ob er at onal eed- h ldren n a l ent la roo. ba to tudent tea her oun elor and other elano 200 o ord Mo Morrell 19 Clients as behavioral self-recorders. Engag- a an 19 2 haneu M ntyre 200 alz n l ent n ob er n and re ord n the r o n ohn on 19. beha or ha be o e n rea n ly o ular. n en youn l ent o e a youn a year add t on to the ad anta e o de rea n or load old ha e u ed el -re ord n Ballard Glynn or re tered behavioral technicians or contin- 19 en h ll ol 19 2 azd n gency managers, self-recording frequently results 19 a hore en Mahoney 19 Bl et in desired behavior change e.. a ato 19. Sel -re ord n not ho e er a le a t e ura 201 d ard 199 u ta ae a a ay r t a ear. ho a 19 ound that e ond- 200 ar on el on aye 19 0 az- rade tudent ar ed n a ura y o re ord n ro d n 19 a ern h ld unla lar e al to 9 er ent. Should your l ent el -re ord 199 nne uer t Ruther ord Glenn 1991 na urately or unrel ably the be t ay to deal th ra el S hlo Al er 199 ll el on that to re n or e a ura y and rel ab l ty per se. 19 2. h ha en e e ally n el - ana e ent hat u ually ol e the roble Broden all ro ra Ma e rato h ll 19 and hen M tt 19 1 en et al. 19 2 azd n 19 a. due to urrent ont n en e the l ent ha a e ted ar ab l ty n re ord n al o ay o ur be au e ntere t n han n h or her o n beha or o - h le o e tar et beha or er t le and a ore-Boy e 19. hu el -re ord n an re e re ord n e.. an er to ult le- ho e be e ed not only a a ean o data olle t on but ue t on nu ber o button buttoned a e r t- al o a a o ble nter ent on. urther ore a e ten roble orre tly ol ed rodu t o leted hall ho later on el -re ord n an nte ral art other are ore a b uou. o are the latter th o el - ana e ent or el - ontrol ro ra. u h er onal ual t e a hy oo erat e Sel -re ord n y te ay ta e ar ou or. neat and other ub e t e de r tor. Su h ter o llu trate eo le an u e a y te a le a need r t to be de r bed n o erat onal lan ua e. tran err n ob e t ro one o et to another hen and date or art at n n el - on tor n ore the re uen y o a art ular beha or on need to be are ully tra ned u er ed and the r a ard or e e o a er or u e a er onal d tal or n are ully al brated a o nt under ored by a tant A or an od. n a t one o u e t hore en and Mahoney a re ord o the nu ber o a e he rote ea h day hen o o n her d ertat on and on n ed ra n n n the d r nat on and re ord n that ere t not or the re n or e ent y elded by the o a beha or e ent al. Su h tra n n ay el -re orded data he ould ne er ha e n hed. be enhan ed by odel n ed ate a u- 150 ra y eedba y te at re n or e ent and ab l t e h le a urately re ord n data on the or- raduated tran er o re ord n re on b l t e re tne o ea h one a e bly o ood e ba and e ternal to el. rete beha or and al o en ur n ea h art ant a ety. le re ord n y te a ear to enhan e Should you be a ed th th red a ent o e el - on tor n a ura y. 19. olut on are o ble. ne alternat e to ob er e and re ord nter ttently at redeter ned rando Se eral re ear her ha e u e ully tra ned t e throu hout the day. hat er t the ob er er nd dual to el -re ord u h beha or a re a n- ont n en y ana er to return to the re ular ta n on-ta ern et al 199 otor t ood unt l the ne t re et t e a ter the br e ob er at on M ltenber er u ley 199 n n er od. oh ood 2001a n rea n re n or er Re ardle o ho re ord n data el -re ord- del ery oo et al. 201 and tutter n a a- n or other re ord n t ortant to a e ure an M ltenber er Arndor er 199. ot only the er on re ord n u ently tra ned. More- d d the r l ent learn rel ably to re ord the e beha - o er t ortant to ut l ze hat e no about or but hen the el -re ord n a o b ned th tra n n hen tea h n nd dual ho to re ord other re n or e ent-ba ed ro edure the roble - beha or. Matthe and a o an 201 o ared at beha or ere u e ully redu ed or elimi- u n d da t tra n n er u a tra n n ba ed u on nated. no n a e do no that art ant beha oral ll tra n n BS.e. n tru t odel beha or o ten ro e a the r el -re ord n ra t e and ro de eedba to tra n ara ro e - be o e ore rel able e an ee hy one u t onal ho to olle t data and de on trated that the be aut ou about a u n that el -re orded data ara ro e onal too ore rel able and al d data a urately re re ent ba el ne er or an e attern. hen they ere tra ned a BS. th that a d t nle a o-ob er er ere urre t t ou ly to re ord important to incorporate systematic training proce- ba el ne data r or to n ol n the l ent n el - dure hen you tra n your ob er er. re ord n depending on self-recorded data alone for establishing a true baseline may be futile. Other a tor ay n uen e the al d ty o el -re ord n Automated Recording Systems a ell n lud n ha n a e ted ntere t n the ten our a e to taff ho e ole or a or ta ro re o the ro ra o a ore-Boy e to ob er e and re ord on o n beha or on- te 19. R hard 19 1 on luded that th nd - l ted. n u h n tan e an alternat e to turn to dual ot ated to el - on tor and han e the r nternet-ba ed deo- or aud o-re ord n that an beha or el -re n or e ent ay o ur naturally be ored.e. treated a er anent rodu t data. 1 o e arat n el - on tor n ro el - and analyzed ub e uently lo ally or ro a d - re n or e ent ay not be o ble. tan e. Su h re ord n al o ha e the ad anta e o er tt n re lay n to n lo n or eed- Contingency managers as data recorders. n u a t on enabl n the u er o t on o t e S e al arran e ent are ne e ary hen on- nd ator and other eature. e ha e u ed deo t n en y ana er are not only re on ble or re ord n to ea ure h ldren o al ntera t on le ent n the ro ra but or al o re ord n tea her a l at on o art ular n tru t onal beha oral data. ubany and Slo ett 19 noted ll h ld eal-t e beha or and any other that la roo tea her are o et e o n ol ed ur o e. ne on derat on thou h the t e t n n tru t onal a t t e that they o et e or et ta e to re e the re ord n and ore and re ord to re ord and a a re ult the r o o te data ay be the data. unrel able. ou an a ne the a e ould ha en ortunately o uter zed re ord n y te with other contingency managers. For instance, it ay o e to the re ue. arly on Bernal G b on ht not be real t to e e t heltered- or ho ll a and e e 19 1 de elo ed a de e or taff to le ent nd dual beha or- nter ent on auto at ally a t at n an aud o re order at ar - lan a ro nd dual th de elo ental d -  8 151 ou t e n the da ly rout ne o a a ly. n a re ent or rea t ty to the re ord n de e to ub de tudy d und et al. 201 u ed a o nt-o - e be ore o ally re ord n and ra h n ba el ne a era to ea ure eye az n n youn h ldren data. A rule o thu b or a u n that beha or ha th aut. be o e rea onably table and no lon er rea t e to Ad an e n o uter te hnolo y ha e led the ob er at onal ond t on to e a ne the data to the de elo ent o ar ou other y te or o nt and a t or a er e onta n n no ne h h re ord n data n lud n u n de to o uter or lo o nt or e eral at lea t three or ore e - la to o uter and o et. ahn and ata sions in a row. 199 ro ded a r t al and n or at e re e o a nu ber o o t are ro ra a a lable or data olle t on ur o e. S n e then th te hnolo y ha REFINING DEFINITIONS ont nued to de elo. or e a le or tho e ore te hnolo ally n l ned or at lea t not te hnolo - arl er e tal ed about the ortan e o de n n ally hallen ed a on and on 200 ha e beha or o erat onally. e only har ly de ned ro ded a ta analy or de elo n a o - en t e ea ure an ro de u th the n or a- uter zed data olle t on y te or a o et. t on ne e ary or al dly re ord n ra h n and d ard 2000 S dener Shaban and arr 200 interpreting behavioral change interventions. al o ha e de r bed the r o uter-a ted ob er- at onal o t are and oldu S n and e elen- Sensitivity of Observational Measures bo h 2001 the r de n or olle t n ana n and re ent n t e- tru tured data ro deota e he ore re ely you de ne your ea ure the and d tal ed a le. Analy e o eb te t ore en t e they ll be. hy th ortant and o the ontent o e a l orre onden e th n really ru al tuat on e ery b t o han e a ro r ate er on o our e are broaden- n the r ht or ron d re t on an really atter. n o ortun t e or ondu t n rel able ea ure- on der or e a le the a ura y th h h a ent o ar ou beha oral la e. n u t one o nur e dole out ll to at ent. ou ould ant to any e a le anaher Bole A er Gordon ea ure er or an e n a re e ay a o ble and Se er on 200 u ed t to a eb te to be au e e en one error ould ean the d fferen e a e art ant e o ure to n or at on about bet een the at ent ro ed health and or o - a o e- ree ub t tute or o n. n one o our ort or n rea ed llne d o ort or e en death. o n eld tud e e an Sulzer-Azaroff le - hen re ord n and ra h n the data you ant to n a ad ro ett 200 e olle ted deo re eal hat ha en n n the o t re e deta l re ord n o arent or tea her to on tor ro a o ble. So n tead o o b n n all the data on a d tan e the er or an e o arent and tea her a ee -by- ee ba you robably ould ant to they ro re ed throu h an nternet tra n n our e loo ore lo ely ay on an e ode-by-e ode and today real-t e ual o un at on y te ba. hat you are a ter are not only h h le el o l e S y e M are be n added to the ABA ob er a- orre t er or an e but al o on ten y be au e t onal re erto re. ater e return to a e other a nd ated e en a n le ar at on n the ron h tor ally ortant and u er- r endly ethod or d re t on ay ean the d fferen e bet een l e and both re ord n and analyz n data. death. ater e return the related to o lott n Re ardle o the ob er at onal y te you data to enable ready a urate analy. hoo e thou h at h out or art ant reactivity, the aty al beha or attern that o ur a a un - t on o the art ant be n ob er ed. Rea t ty SCHEDULING BEHAVIORAL al o ay be a roble th ntru e or o auto- RECORDING at re ord n u h a u n handheld a era. le ntru e ob er at onal y te are una a l- Al ay ee n the on e t o validity that the able be ure to allo u ent t e or adaptation ethod ea ure hat t u o ed to ea ure n 152 nd you need to hedule your beha oral re ord- hether and o ho there ha been a d ern- n e on n u h a ay that the data you olle t ible han e. or n tan e you ay ant to o are truly ea ure hat they are u o ed to ea ure. the data re ealed under ba el ne er u tho e th n t e o day day o the ee eather ond t on one treat ent ond t on or tho e enerated n other ta a n ent hy al or o al urround n treat ent er u t ll other and o on. er onnel the tatu o the art ant n ter o at at on or de r at on re ard n ood u d a b - ent eather or te erature ond t on and a yr ad Definitional Detail o other ar able threaten to n uen e the data you o deta led do you ant your er or an e ea- need to a e erta n the e are e enly d tr buted ure to be our e that de end on the a o throu hout ba el ne nter ent on and ollo -u your ro ra. t ly to n rea e the rate ha e. or ob ou ly they are not o ontrolled o an already o erat onally-de ned and al dated you ll not no h h one or o b nat on a ea ured er or an e u h a nu ber o te re on ble or any han e ta n la e dur n rodu ed to tandard nu er al roble ol ed your beha oral nter ent on ha e. orre tly or a ety re aut on ra t ed all you need do to de de yes or no. ount n u n Frequency of Recording and ra h n the data robably ould be u ent. n other r u tan e you ay need to return to ther ue t on ar e at th o nt o frequently and re e your ob e t e unt l anyone th the hould e ea ure under hat ond t on A a ro r ate tra n n and u er on an rel ably you ll ee later on e u t re e our ob e t e a e the er or an e. on e a a n. e are on erned ly th n rea - n or de rea n the rate o a beha or a n nu - ber o rodu t eet n or not eet n tandard or Unit of Analysis orre t and or n orre t an er on a te t o er en hat ze hould you ele t to hoo e a the un t to er od o t e then all e need to do to u e our be analyzed he un t o analy or a e n al dly-de ned ob e t e a ro tandard t e er - and e a n n er or an e ar ab l ty ht be od. n tead our a to ha e ne beha or or lar e or a total rodu t on ta. a e the e a le tea h our l entele to d fferent ate one t ulu ro o ern ho a ood e ba n a heltered another e ay need to ob er e and re ord a o ten or ho. n her a e you ht de de to ob er e a o ent-to- o ent or tr al-by-tr al re ult. lo ely and re ord the t e t too her to ll ea h ba o er t e be au e that ould er t you to Duration of Recording Phases dete t and re n or e e uen e o re on e o ur- r n at h her rate than u ual. o lon duration e a nta n ba el ne or ea h Al o a your el and your a o ate ho nter ent on ha e de end on the data the el e. refined you ant your analy to be to hat e tent ne o e eral rule o thu b de end n on other you ant to o b ne data or ur o e o ra h- a tor u h a ur en y to le ent nter en- n and analyz n. ou ay de de that ra h n t on ra t al on derat on and o on to con- every single response, as in cumulative re ord n tinue to assess performance until the data become ll ro de you th the n or at on you need. stable according to your pre-specified definition r erha you are on der n o b n n your of stability n e ample of such a definition is that data by nter al o t e or rou o re on e. data reveal no new highs or lows for five (or some An e a le o the or er ould be er n et larger number) sessions in a row. Stable ba el ne o te a e bled on the ba o day o the ee ro de a tandard a a n t h h to e aluate ub e- ee ure.1. ot e that our art ant ern uent han e n le el trend or ar ab l ty. ou a e bled bet een 2 and ood e ba er ant to be able to e a er e o data o nt to ee day. a ne tho e data ore lo ely. o you  8 153 Number of Goodie Bags Fern Asseembled by Day of the Week over Seven Weeks 700 600 500 Number Assembled 400 300 200 100 0 Mon. Tu. Wed. Th. Fri. Day of Week Figure 8.1 Ba ern a e bled by day o the ee not e anyth n e al about the you a d he ar ou ay o lott n and analyz n data that ern rodu t ty a enerally lo er on r - that e ha e u t re ented ro de u th o e day and al o t a oor on Monday you ould ntere t n hy othe e about a tor that n uen e be orre t. you ere her u er or hat ur- ern rodu t ty. et there are any other le el ther n or at on ht you ee hat han e o data analy. or n tan e you ould e a ne any ould you on der a n n her ee ly the nu ber on the ba o hour o the day or n- hedule r the e data ht re you th ute th n the hour. r you ould loo at eather the otent al alue o n e t at n hat ha en ond t on or day o the onth or ern hy al to n uen e Monday and r day lo er ore. ond t on. or n tan e doe he et to bed too late on Sun- Su o e you anted n tead to e a ne ern day he t red by r day or he d tra ted by ee ly rodu t ty ore lo ely a n ure.. ee end lan ot e the b ur r e dur n ee e en Fern Su o e n tead you ere to o b ne the data rodu t ty a aty ally h h on Monday and nto ee ly data or the nu ber o hour t ta e to hur day o that ee. h n or at on ht end a e ble a en nu ber o te. ou then ould you off on a on to d o er hat e al e ent e a ne the ra h re re entat on o the data a ay ha e o urred dur n that ee e ther th n n ure.2 to ee any art ular re on e on- or out de o the or ho. Maybe you ould learn urat on e er e that you then ould rodu t ely that on tho e t o day ern ould u e her earn n analyze nter ret and ub e t to ar ou nter en- to lay b n o o no you ll ha e ound a oten- t on unt l you ound one that a o t effe t e. t ally aluable re n or er or ern. Alternat ely you ay deter ne that ra h n Maybe n tead you de ded to o b ne the ea h re on e a done n event or cumulative data nto b - ee ly data or the nu ber o hour re ord n de r bed belo ll ro de you th t ta e to a e ble a en nu ber o te. ou the n or at on you need. then ould e a ne the ra h re re entat on o 154 700 600 500 400 Number 300 200 100 0 Mon. Tu. Wed. Th. Fri. Days of Week Figure 8.2 o bnn ern data by day-o -the- ee a ro a e en- ee er od 95 90 85 Tues 80 Wed. Thur. Number 75 Fri. 70 Mon. 65 60 55 50 1 2 3 4 5 6 7 Weeks Figure 8.3 ern ee ly rodu t on  8 155 tho e data to ee any art ular re on e attern or whether the behavioral pattern is becoming more e er e attern you ould then analyze nter ret variable or consistent. and rodu t ely ub e t the ro ra to ar ou By on ent on ABA ra t t oner and re ear h- nter ent on unt l you ound the one that a o t er tend to u e a tandard ra h or at. Be o n effe t e. ll ul at e n and nter ret n a ra h re u re a b t o read ly- a ned te hn al ll. ere e u de you throu h the ro e by re e n the o o- GRAPHING BEHAVIORAL DATA3 nent o an ABA ra h. A you already re o n ze a beha oral ra h ro- de a ual re re entat on o beha or o er t e. Standard Format for a Graph By lott n and onne t n data o nt e be n to not e er or an e attern e er n. he on u- eue ure. to llu trate n l ed or rat on or ed by re eated ea ure o the beha - e eral eature o a a rly tandard ABA l ne ra h. or a t ty ally o ur that ta en be ore any A line graph the ty e o ra h de r bed belo and nter ent on aybe u ed a a tandard or baseline u ed o t re uently n th eld u ually displays a a n t h h to a e any ub e uent han e data that can be scaled along some dimension, such that ta e la e e ther a t e a e and or a ne as time or the order of responses in a sequence. The ond t on are ntrodu ed. A nu ber o ra h n ra h n ure. d lay ho the el - n ur ou on ent on ha e o e to be a e ted dur n the beha or S B o a h ld na ed oy han ed o er e olut on o the eld o a l ed beha or analy. u e e ond t on ur n ba el ne B SB o urred relat ely o ten but de rea ed to ery lo le el dur n an nter ent on ro edure alled R Graphing Guidelines non ont n ent re n or e ent. a ne the deta l o on tru t n a ra h o th ty e A ter beha or re orded a e ha e een the re ult n n or at on u t be ra hed ed - r t are the ert al and hor zontal l ne he ately to er t han e attern to be ob er ed and vertical l ne alled the y-axis. The ordinate do u ented. Gra h re re entat on o data are the the y alue. The y axis usually provides a r ary ba e or on tor n hether y te at standard for measuring the dependent vari- alterat on n the beha or o on ern o- ary able (the behavior u h a re uen y th en beha or- han e ro ra.e. hether e u ber er enta e etc., and hen the beha or- han e ro ra le ented contains axis scales or numerals. n ure the beha or han e orre ond n ly and or. the nu eral on the ord nate re re ent- de d n hether any ro edural han e are nd - n the y- alue be n at zero and ro re ated. By ually n e t n the ra h e an n un t o e. a data et onta n any deter ne hether the er or an e ea ure and zero o nt ay three or ore or the a e or t trend and or t de ree o ar ab l ty y te - o lar ty the y-a an tart a l ttle belo atically are changing while the intervention is in zero rather than at zero a n th e a le. la e. o e then deter ne any han e are he hor zontal x-axis, displays the label n the ant ated or the o o te d re t on hether for the observational sessions. These may alterat on n trend are a elerat n or de elerat n be o o ed o e ob er at onal e - on u ually de r bed n ore deta l n the 3 or reat n n le- ub e t de n ra h th M ro o t a o any n narrat e or tandard un t o Excel M 200 ee on et al. 2009. Al o ee Barton t e u h a hour day a n th e a le Re ho 2012 or ra h n data th M ro o t e weeks, or months. he r t data o nt ra hed 2007 M, 2010 M and Ma M2008 & 2011. eo hand 201 de r be ho to add ha e l ne and the r re e t e label all on ure. then, has an ordinate alue o o h h tay e bedded n the ra h hen they are u dated or o ed hen u n M ro o t el M. 156 ABAB BL NCR BL NCR 40 35 30 Frequency of SIB 25 20 15 10 Joy 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Day Figure 8.4 y al thdra al4 A-B-A-B ra h 20 o urren e o el - n ur ou beha or on Steps in Graph Construction day one noted on the abscissa the - alue. ra the ert al y-a l ne 1. e t a ter allo n or ada tat on e add the ra the hor zontal -a l ne 2. data points a they are olle ted. ually the r t e eral data o nt on t tute the baseline. A ter nd ate the label and un t o ea ure ent u ent ba el ne data ha e been olle ted labeled on the a e. and lotted on the ra h o e ort o han e trat- Add the data o nt and y lot repre- e y the intervention or treatment u ually ntro- senting the responses across time. ote du ed. o nd ate hen that ha en e add and that you onne t data o nt within a phase label ert al l ne and label ea h ha e e.. B and disconnect them between each phase. R bet een the la t ob er at onal e on and Add vertical phase-change lines. ha e tho e enerated under the ne ond t on. A you l ne an be ol d a n ure. or bro- ll learn n ore deta l hortly a u n er or- en a n ure.12. These indicate that a an e attern ha e han ed n the ant ated new condition is in effect; usually a differ- d re t on e then e ther thdra tho e ond t on ent “treatment” or independent variable is to ee beha or return to t or er attern n being implemented. For example, a phase- the A-B-A-B de n a ly the han e trate y to han e l ne ould be n erted between the a ne beha or er on or lo at on n the ult le- la t ba el ne e on and the r t treat ent ba el ne de n or ollo a d fferent e er ental e on e.. bet een ba el ne and R de n ath o e o h h ll be o ered n ha - non ont n ent re n or e ent. ter 2. ere then a re e o the te you ollo n on tru t n an ABA ra h u n ure. a an e a le 4 re er to t h n the ont n en e o on e uen e no are the o o te ro hat they had been re ou ly. n the re ent 4 So et e the ter re er al u ed to de r be th de n. a e that ould ean h t n to ro d n re n or e ent e hn ally thou h a you ll ee n ha ter 10 re er al contingent on the response.  8 157 Add a phase label. he ha e label uses ore than one data et be ure to u e d - a simple descriptor of the condition(s) in erent y bol and or onne t n l ne or place e.. ba el ne treat ent ollo -u ea h ee ure. and.. et.. G e the ra h a br e de r t e label. he t lar e area o dead a e by u n a label hould a e t o ble or another brea n the a. a e are thou h not to er on to e a ne the ra h and under tand d tort the a n tude o effe t hen you hat ha ened hen and under hat ond - do th. ot e on ure. belo that tions. the han e ro day to day 10 a tually only about 1 er ent. ote that the nu er- al on the ert al a y-a u ro Graphic Variations 0 to 0 then ont nue n n re ent o 10. y ally e lot only one et o data on a ra h. needed or lar ty hould any three o e er lott n both data et on the a e ra h or ore data o nt all on the zero l ne an be e e ally n or at e art ularly hen e tend the ord nate to belo zero. ult le et o d fferent nd o data are e e ted the ar at on n the data lot ee to relate to one another in some way, as in the case to ob ure a o ally ortant han e o re a n n on- and off-ta and nu ber o rob- e and the ord nate. le o leted. Should you de de to do o eth n ually only one data et lotted on a lar you an u e both the le t and r ht y-a e. ra h. o e er hould you de de to lot he le t ht re re ent off-ta the r ht rob- Percentage On-Task 100 90 80 Percent Percent 70 60 Baseline Intervention 50 0 1 2 3 4 5 6 7 8 9 10 Days Figure 8.5 Brea n the ord nate hen no data a ear n a ran e ot e the brea on the ord nate bet een zero and ty.  158 100 BL DRA 40 35 80 Off-task 30 Math Problems % Interval Off-task 25 # Completed 60 20 40 15 10 20 5 0 0 1 2 3 4 5 6 7 8 Sessions Figure 8.6 Mult le data- et lot double ra h le o leted. A the ra h be n to e tend n u h n or at on ba ed on data o b ned a ro t e you an ee and ho the han e relate to e on hour or e en day or ee. one another. you de de to lot ult le data et ure. d lay erry ho e hen dur- on a n le ra h u e d t n t y bol and a o - n the r t tr al he er tted to ele t one te ated onne t n l ne or ea h. n ure. lo ed ro a tray onta n n the ull array o the ood uare re re ent off-ta beha or h le lo ed tr - he re orted to l e n lud n a oo e a h a an le re re ent ath roble o leted. e e o andy a l e o an a le or a ear. he oo e n hand do n. ur n the e ond et o ho e the oo e ha been re o ed ro the array Bar Graphs and he hoo e h t e andy t e and the other two once each. By the time the assessment is Bar graphs are more appropriate than line graphs o leted e no that erry l e all the ho en hen there no on en ent d en on alon h h te but hen e really ho e to motivate h t to arran e the data. hey ro de a b rd eye e o ould be e to ha e a oodly u ly o h a or- o e a e t o an nd dual beha or u h a the te oo e and h on hand. ho e he enerally a e or the n rea e n ro- en y or rate o a er on learn n a ll. nl e line graphs, bar graphs do not represent response Cumulative Records sequences. Rather, they are used to compare dis- crete sets of data that may be, but are not necessarily e the ebb and o o the a e and the t de related to one another; or to summarize performance beha or h t o er t e. he ont n ent rela- within a condition or group of individuals. Typically t on n effe t at one o nt ay han e at another. hn o your o n n n attent on de end n  8 159 Figure 8.7 re eren e a e ent ra h bar ra h er enta e ele ted th n tr al th that te ot e al o the redu t on n ar ab l ty o the data. tan e robe ere ta en ro another anta e o nt orn n robe at a d ffer- ent t e o day. on e uen e o nternal and e ternal e ent. h le ulu -re on e relat on ore lo ely. arly on n the at h n tele on your to a h ro l and you ABA eld n e t ator a re ated th a t nd- a ne the ontent o your re r erator. our ell n ar ou tool or u h ur o e. he e n luded hone r n ed ately o and n your atten- the ult le-e ent re order h h er tted e eral t on h le d tra t n you ro your read n. our ar able to be ea ured ultaneou ly re on e lo e o an on loo an ry you nterru t your rate re entat on o d r nat e and re n or - eal re arat on to try to nd the au e. h le n n t ul and o on the u ulat e re order and e eryday l e the e te orary h t u ually aren t the lo ar th hart. e de r be tho e ne t. trouble o e but o et e they an be. e orar- B.. S nner 19 n ented a de e alled ly thdra n attent on ro a ob o erat on or the cumulative recorder that rodu ed u ulat e ro the road h le dr n to an er the ell hone data lot. he e lot er tted n e t ator to ay lead to an n ur ou a dent. A h ld thro - n e t o ent-to- o ent han e n rate d re tly n a tantru n the d t o a nor ally al t an e tre ely u e ul etr. n the u ulat e to the u er ar et or a re taurant ay be e tre ely re order or nal laboratory a l at on ea h o d turb n to you and other. n u h a e you ay the or an ra dly re eated re on e de e ted want to inspect behavioral patterns more closely, a en u ard by a ed a ount a the re ord n ro o ent to o ent rather than ro e on a er o ed at a teady eed. h rodu ed an to e on or day to day. o erall lo e or ur e lar to the deal zed a e n tead o n e t n data o b ned th n and llu trated n ure.. n a tual u e the he ht or a ro e on then e re o n ze that there are o the en de e t on u h aller allo n or t e hen t hel ul to be able to e a ne t - any any ore re on e be ore the en rea he 160 Figure 8.8 A l ed de r t on o the ay a u ulat e re order o erate. ro the ly hn e on eb te © Dr. C.A.P. Kenyon. The paper rolls and each time a response occurs, the response pen moves upward until it reaches a particular height. At that point, the pen drops back to its lowest point and again begins its journey upward once again. Delivery of reinforcers is indicated by the small hatch marks. Hatch marks on the lower pen indicate some particular event, such as a change in illumination. the top of the paper, then dropping back to the bot- behavior, such as drinking? Maybe the ambience tom to again begin its upward climb.) of the laboratory altered, as in a nearby female of The paper rolls and each time a response occurs, the a e e e n heat or the l ht turned off or the response pen moves upward until it reaches a someone slamming the door? Many a new experi- particular height. At that point, the pen drops back to mental direction has been the outgrowth of unex- its lowest point and again begins its journey upward e ted e ent be au e the r effe t a re e ted on once again. Delivery of reinforcers is indicated by the cumulative record. the small hatch marks. Hatch marks on the lower Cumulative records display rates of some pen indicate some particular event, such as a change event (usually a response), in the form of changes in illumination. in the slope or curve of the record (e.g., number of The cumulative record has the advantage of re on e a e t e er od a a un t on o revealing, at a glance, the frequency of responses ond t on n effe t. he re ord ho ho ra dly over time, or rate. The researcher can examine the or slowly the responses are repeated. When you record to note places where the rate slows, quickens, view a cumulative graph, you should recognize that or becomes especially variable, and match those the steeper the slope, the more rapid the response changes with other events. Did the rate diminish being recorded. When the data line goes sideways, gradually over time? Perhaps the subject (or partici- or straight across, it means that no occurrence of the pant) was becoming sated with the reinforcer. If the behavior was recorded. Additionally, the cumulative rate varied, was that because other response options record permits one to view at a glance the total num- and/or reinforcers became available or because con- ber of responses emitted (accumulated) during the suming food changed the probability of the other recording period.  8 161 n her do toral d ertat on re ear h one o the 8.9. Su o e on ay 1 Mar aret ran le and author 5 u ed a u ulat e re order to a e the on ay 2 he ran le. he r t data o nt lot- re n or n un t on o d fferent la e o oten- ted at 1 - nter e t on the e ond data o nt t ally re n or n t ul andy d lay o olor- at the 2 nter e t on be au e our le ere ul a e e uent al nu eral and o on on the added to the three o the r t day. n o e day at h-to- a le beha or o ty ally- un t on n e.. bet een ay 2 and and ay and the l ne h ldren and tho e th de elo ental delay. re a n at be au e Mar aret d d not run on tho e Be de d lay n the total re on e re uen y at day. A u e ul eature that by loo n at the la t a lan e the u ulat e re ord er tted u to ee data o nt e an ee ed ately that Mar aret ran han e n rate ro one e er ental ond t on a total o 29 le o er the t el e-day er od. to another. he tee er the an le de r bed on the Re o n ze then that u ulat e re ord allo u ulat e re ord the ore ra dly the h ld a u to de on trate lo al a ell a o erall re on e re ond n. oday a l ed beha or analy t u e rate. enrod alla e and yer 200 r t u ed ore ad an ed n tru entat on to rodu e u ula- le l ne ra h to e a ne d fferen e n le - t e re ord but under any r u tan e the ut l- el o er or an e under t o ond t on h h- ty o the u ulat e re ord re a n. re erred and lo - re erred re n or er Be au e they d lay both the rate o re ond- under a progressive ratio ( R) schedule n a R n and total nu ber o re on e lo -te h u ula- the ratio of required responses to reinforcers con- t e re ord al o an be u e ul or a l ed beha or tinues to increase. But the rea on or tho e d ffer- analy t or n n the eld a llu trated n ure en e a un lear. ere they due to d fferen e n the effe t ene o the re n or er or the art ant 5 Sulzer B. 19. Mat h-to- a le beha or by nor al and a on lu on un u orted by a tudy o re ou retarded h ldren under d fferent re n or n ond t on. er or an e attern under ed rat o hedule n ubl hed do toral d ertat on n er ty o M nne ota. 30 25 20 # of Miles 15 10 5 Margaret 0 1 2 3 4 5 6 7 8 9 10 11 12 Days Figure 8.9 u ulat e ra h o the le Mar aret ran er day t t ou data 162 or was something else going on within the ses- han e by a a tor o 10 ro.1 to 1 ro 1 to on u ulat e re ord ub e uently ere u ed 10 ro 10 to 100 and o on. he ele ant eature to de on trate d t n t e th n- e on attern o o u h a ra h that hene er beha or han e re ond n ee ure.10. or e a le Sa by o e a tor e.. a doubl n th n a en art ant 1 u ulat e-re ord data re ealed that t e er od the lo e o the han e loo the a e he on tently re onded dur n the R- ond - hether you tart th a ery h h or a ery lo le el t on a tee lo e but he too ore brea dur n o beha or. a ne n rea n one h ld a urate the R- ond t on the lo e a atter e en re ond n ro 2 to orre t an er er day and thou h he or ed at about rou hly the a e rate or another ro 10 to 20. you lotted the e on tan- both re n or er under a ure ed-rat o hedule. dard l near ra h the r t h ld ro e ent o u ulat e re ord n al o er t u to e only 2 orre t an er that day ould ee all the o nt at h h re ond n ea e the brea o ared the e ond h ld ro e ent o 10. o nt a the hedule re u re ent n rea e But n ter o er enta e both h ldren ro ed de and n ore and ore re on e or a ter and by 100 a a re ult o your tea h n. he r e ual ty a ter re on e rate. Su h n or at on n or n ter o er enta e han e ould ho a e ual u about ho to ele t re n or er and hedule ro re lotted on a tandard elerat on hart. re u re ent the h her the brea o nt the ore ne a l at on o beha or analy that rel e otent e a u e the re n or er to be e e ally hea ly on u n tandard elerat on hart re- the oal or the art ant to u ta n re ond n cision teaching,6 an nd dual zed n tru t onal at a rea onable and on tent rate. So hen Ser e ethod that e ha ze rate bu ld n and hart n belo ed ae tro radually n rea e the t e he a o er or an e nd ley 1991. ot e on ure Ser e to ra t e the ola to re e e re n or e ent.11 ho the ra h a o odate re on e rate o n the or o ra e Ser e ont nue to e and h 001. er nute about a n le re on e n a day to ra t e hedule unt l he rea he h l t o e 1 er nute and o on to 1000 er nute. nd ley hour er day the brea o nt be au e he u t u ed a e -lo ar th ale or the y-a o that be n r t n a ter a er or h zoolo y our e. art ular beha or that o urred o er ery d ffer- hat re tan e to han e n or u that the re n- ent t e ale ould e ently be o ared to one or er o er ul ndeed Gra e e n 199. another. he hart n luded re uen e ran n a dely a ro 1 er day to 1 000 er nute on a n le ra h o er n 1 0 u e e alendar day Standard Celeration Charts about one hool e e ter. hat ade t o ble to lot both re uent and n re uent beha or on Another ay to ho nu ber or re uen y o the a e ra h o er our onth or ore. re on e o er t e to u e a tandard elerat on hen ht you nd t ra t al to u e a e - chart ure.11. en h ldren an learn to u e lo ar th ra h r ar ly hen your on ern th ra h n te hn ue but t ta e a e tra n n ro ot n rate o re ond n o eth n you ll sessions. The standard celeration chart is a varia- ant to on ern your el th be au e t a a t tion of a semi-logarithmic chart, which shows pro- that the ore ra d and uent the rate o orre t portional or relative instead of absolute changes in re ond n the ore durable the learn n ohn on behavior. Stated ly unl e the equal-interval ayn 199. or e a le t o tudent ea h all nter al are the a e ze l ne bar u ula- ore 100 er ent on an e a. o e er one tudent t e and other graphs, the semi-logarithmic scale too only 1 nute to o lete the e a the other allo data to be ueezed nto ro re ely t hter too 0 nute. he r ore re orted n er ent do and t hter bundle. n a tandard elerat on hart not nd ate ho ht ha e a tered the ater al han e o e ual ro ort on are ho n by e ual ore thorou hly. o e er be au e elerat on hart- d tan e on the ert al a e u n a e -lo ar th- mic axes scale with 6 to 10 cycles on that y-axis. he a or y-a d on n ure.11 all ho 6 See ha ter 1 or a ore deta led de r t on o re on teaching.  8 163 PR-HP Cumulative responses 1200 1000 800 PR-LP 600 400 200 Sam 0 0 1 2 3 4 5 6 7 Sessions Cumulative responses 150 PR-HP 125 PR-LP 100 75 50 25 Cedar 0 0 1 2 3 4 5 6 Sessions Cumulative responses 250 PR-HP 200 150 100 PR-LP 50 Aden 0 0 1 2 3 4 5 6 Sessions PR-LP Cumulative responses 700 600 PR-HP 500 400 300 200 100 Elliot 0 0 1 2 3 4 5 6 Sessions Figure 8.10 A e n re n or er oten y by ean o u ulat e re ord n. ro enrod alla e yer 200 ure. © 200 by the So ety or the er ental Analy o Beha or n. 164 Figure 8.11 Standard celeration chart ing graphs rate performance, it would show the dif- a nta n n ade uate ea ure ent r ter a. But a ference between the two students. Semi-log graphs, ohn on and ayn 199 ha e re orted re ard n then, accommodate the kinds of major changes we the r Morn n de Model o Generat e n tru t on ho e to ee n uen y bu ld n.e. eed n u following brief training, students can learn to assess, rates of answering math problems, writing spelling graph, and input data on their own performance quite ord an er n ue t on orre tly n utt n ey- a urately. board entr e and o on. Add t onally the tandard celeration chart lends itself to comparisons across Monitoring Changes in Behavior any nd o beha or th n ar ed onte t. or e a le tudent a ade ro re ere re orded Patterns nd dually on tandard elerat on hart n d fferent So et e e en hen e do not dete t a han e hool or hool y te t ould be ery ea y to n le el or a trend e ay ob er e an ortant see which ones were generating the most progress han e n ar ab l ty. on der the alue o atte t- n ter o er ent ro e ent. But u h re ord ing to increase the level of a response such as pro- cannot be produced unless adequate resources are ot n h ldren rate o ear n a ety hel et de oted not only to the re ord ee n but al o to  8 165 and do n o ore on tently le variably h le der hone r n n a arental tone o ur b n. an outen, an outen and Malen ant r or to a re on e or a ter a le lo o 200 a o l hed that by u n peer goal-set- an ob e t or attent on. t n ubl o t n o the er enta e o orre t hel- eter ne beha oral un t on a d re t et u e and hared re n or er. e ure 8.12 a e ent a n e ally re ent n or and not e ho u e ul they ere n a h e n thdra n attent on ollo n ry n. both ur o e. ete t th n- e on trend rate o ey n and o the el e ar able data are not ne - entry dur n a da ly ob er at on. e ar ly roble at. e erthele a u n your objective is to promote behavioral increases you Sa e uard l ent u h a not n n tan e an on tor at h and re n or e the beha or o the l ent n e t n arette butt. or e uen e o beha or rodu n the h he t reate u ent data or analy a e - o nt. hen lower rates are hat you are a ter u t n an o erator o ture h le or n at do the o o te Re er e re n or e ent or tho e the o uter nute by nute rather than e uen e o re on e that are een to all at the a l n by day. lower ranges. eter ne ob er at on- e on durat on the A ah e and anley 200 a no led e beha or u h a lou h n ton ue thru t- o b ned data o ten are ea er to ra h e a - n or deburn tu n ha en o o ten ne and analyze or ur o e o dete t n han e that a br e a le ade uately re re enta- n relat on to the ond t on n effe t. hey er- tive. t u to un o er ad u t ent n re on e attern lar y ounter ntu t e re on e attern o er t e n lud n change in level is the rate or ho ould e e t a l ent to be o e frequency or accuracy of the behavior increasing, n rea n ly a re e a ter re e n decreasing or remaining the same? trend is the her ty ally a ored re n or er a a a- rate changing in any particular direction—slowing z ne n that a e you ould e a ne the down or speeding up—over time? and variability o ent-to- o ent data to try to dete t the does the rate of the behavior fluctuate a lot, a little n rea e n the l ent rate and nten ty o or hardly ever?. he e three ter are d u ed n aggression and ear h or other ond t on ore deta l belo. that ay ha e ntruded at tho e t e to et et the bene t o ee n the a e ro a ar her off e.. he la ed her la e or the exact a cost, in that we may risk overlooking essen- l ht le el han ed. t al deta l tho e ubtle but r t al a tor that ay be n uen n o ent-to- o ent beha or. nder tand beha oral ro e e deter n- So hen the ne l able ha en and you ant to n hether the han e a a un t on o unra el the y ter ou au e un t onally related at at on any any re n or er o er and a tor you ll need to n e t ate the tuat on beyond the h tor al nu ber or that ar- ore lo ely and loo at nd dual and aller t ant or de r at on be au e re n or er a re ate o t ulu and re on e relat on to u ent to a nta n the beha or ere d o er hat n uen n the out o e ed - una a lable. ately ollo n dur n and or r or to ea h n le re et t on o the re on e o ntere t. ah e and anley 200. 2 2 ha e l ted and llu - INTERPRETING BEHAVIORAL trated u h r u tan e. e o b ne tho e te DATA th o e o our o n. a ne r u tan e ore lo ely to Re e ber that n and o the el e data or ra h re re entat on o beha or are but a hort- e r be naturally o urr n beha oral hand ean or on ey n n or at on. hey a n relat on hether a art ular e ent thun- the r alue a ord n to ho u e ul they are n hel - 166 Baseline Helmet Program 100 Percentage Correct Helmet Use 90 80 70 60 50 40 30 20 10 Bonita Springs Middle School 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 100 Percentage Correct Helmet Use 90 80 70 60 50 40 30 20 10 Riviera Middle School 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 100 90 80 70 60 Morning Probe 50 Distance Probe 40 30 20 10 Meadowlawn Middle School 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Daily Sessions Figure 8.12 An llu trat on o han e n er enta e o orre t hel et u e. ro an outen Malen ant 200. © 200 by the So ety or the er ental Analy o Beha or n. ot e al o the redu t on n ar ab l ty o the data. tan e robe ere ta en rom another vantage point; morning robe at a d fferent t e o day.  8 167 n u to a e ound ean n ul de on a n alue d ded by the total nu ber o o urren e hether or hen to n t tute a art ular ro ra o o a et o data; or mode, the alue o the data o nt tea h n or ana e ent and or hen to han e o t hea ly re re ented; or median, the alue it. n other ord a o nted out n the re ou o the ore n the ddle o the array o ore ha ter ra h re re entat on need to ontr bute arran ed ro lo e t to h he t. Level changes are to the treat ent effe t ene the treat ent ut l ty ea ly do u ented by ta n the a era e alue o o a e ent. ea h ond t on and o ar n the. Re er ba to To make the most o our data e need r t to ure.. ote the han e n le el de on trated e a ne the lo ely to dete t hat they are or are between the ba el ne and the RA d fferent al re n- not tell n u. Ar ed th that n or at on e then or e ent o alternat e beha or ond t on n both an eel ore on dent about ont nu n or alter n the nu ber o ath roble o leted ro an our ra t e ro ded e ee on tor n the data average or mean M 1. to an M. h le attern a ter any ro edural han e. the er enta e o nter al off-ta d n hed ro o nter ret the data you u t a an M 9. to an M. A trend be o e a arent hen data o nt How much ha the beha or han ed e ee to be rodu n attern that e ther r e or all d o er th by loo n at level changes. o er t e. ure.1 d lay a tual e ual- nter al In what direction or better or or e the data ho n the n rea n attendan e trend at the han e headed? e an deter ne th by annual eet n o the Berkshire Association for looking at trend. Behavior Analysis and Therapy over the ten years o consistent re ularly or rre ularly is ro 199 to 200. e er read ly an ee the the han e that ta n la e n th a e trend that the nu ber o eo le attend n ha been we examine variability. n rea n tead ly ro year to year. hen data are ore ar able an ea y ay A change in level de on trated hen the to de r be a trend to u e your o uter data data a era e alue han e mean or u o the management program to rodu e a le trend- Attendance at BABAT over the past Ten Years 700 600 Number of Attendees 500 400 300 200 100 0 1 2 3 4 5 6 7 8 9 10 11 Years Figure 8.13 An e a le o an n rea n attendan e trend at the eet n o he Ber h re A o at on or Beha or Analy and hera y ro 199 to 200. 168 line.7 he trendl ne an tell you hether the data Variability re er to the e tent to h h data enerated are n rea n or de rea n o er t e boun e around on the ra h. n a l ed beha or and an e you a standard of reference for deter- analy o our e that re er to ea ure o an mining if the rate of change is accelerating or decel- nd dual on o n er or an e dur n ba el ne erating over time. oo at ure.1 orthu and or dur n one or ore nter ent on or ollo -u oll er Serrett 199. and you an ee er od. hen e n u re about ar ab l ty e a e eral trend ate or e o ubl at on to ro the beha or o ntere t be n e tted a rly on- the Journal of Applied Behavior Analysis. ote or tently homogeneously or u n around ro e a le that the nu ber o a er on ll a u - one data o nt to the ne t heterogeneously. hy t on and on beha oral e e e a on eo le n or at on on ar ab l ty o ortant hat th de elo ental d ab l t e n rea ed o er the de end on your ur o e n re ord n n the r t r t 2 year h le other la e o to ho ed la e ho t er e your a h e n the ob e t e o de l n n trend.8 the ro ra. o llu trate o et e ar ab l ty an be roble at a n the a e o a nta n n a e 7 See htt o e. ro o t. o en u hel ond t on. At other t e t an be an ad anta e. A a a e n o nt u o e you are atte t n to shape 8 tt n l ne to an array o data o nt on e -lo ar th a er or larly althou h be au e the ale han e ro one le el to the ne t the l ne that ould be t t the d tr but on te t ha e been de ned to enable on u er o the n or at on o data o nt are ore a t to or ur e. ar ou tat t al to deter ne the n an e o the re ult obta ned. 60 DD skill acquisition 30 other academic other language Percent of Articles 0 68 72 77 82 87 92 60 DD behavioral excess other child behavioral excess 30 0 68 72 77 82 87 92 Figure 8.14 llu trat e data trend. ro orthu oll er Serrett 199. © 199 by the So ety or the er - ental Analy o Beha or n.  8 169 a ne beha or by re n or n only tho e re on e are e ent al ele ent o a l ed beha or analy. that ore and ore lo ely a roa h a oal u - al dly and rel ably ob er n and re ord n data e ely a ro ate the te hn al ter ee er t u to a o l h a nu ber o ur o e ha ter 1. ou ould ant to re ord re ond n to deter ne ob e t ely hat an nd dual or ery o ten ay e ery 1 e ond to allo you to rou er or an e loo l e r or to any nter- dete t and treat the l hte t beha oral ar at on n ent on that dur n ba el ne a ell a dur n the re erred d re t on. onder ho lo ely you ay and ollo n the treat ent or nter ent on. h to e a ne your er e o data o nt or nd - n th ha ter e ha e de r bed e eral o a at on o ar ab l ty or to dete t ea ure and treat nu ber o ho e a a lable to you hen you de n re n or e tho e re on e that nd ate han e n your ra h n ethod who to involve in the pro- en d re t on. e and how you ht re are the ll ully n a nta n n a ety ar at on n re ond- and ob e t ely to d lay and analyze beha oral n an be o e atter o l e and death a n the data. ou ha e e ed le l ne ra h that er- a e o h ldren u n hel et hen r d n b y le t e er to dete t han e n the le el trend d u ed earl er ure.12 an outen et al., and ar ab l ty o the data they ha e olle ted and 200. ho e n e t ator heduled the r ea ure- ha e een o e e a le o d fferent ethod or ent nter al to er t dete t n u h ar at on de r b n and nter ret n re on e attern. and al o to d fferent ally re n or e ro e ent A a beha or analy t you ha e a re on b l- n er or an e. ot e or n tan e that the data ty to olle t and ra h ally d lay data u n ho not only an ro e ent n hel et u e but the ethod de r bed n th ha ter. ou ll u e also in consistency. the e ra h d lay or de on and re o en- o e er o et e beha or u t h hly dat on or beha or han e ro ra de elo ent. ar able. ne u h beha or hy al a t ty. n ou al o ll u e your ra h d lay to e la n a re ent tudy albuena M ller Sa aha and M lt- n under tandable lan ua e to your l ent and are- enber er 201 re ent e en trate e to ut l ze ta er the r ro re and re ult. n add t on be ore hen beha or ar able and n lude ha e ean any nter ent on or a e ent ondu ted you and ed al l ne da ly a era e er ee ee ly u t ha e your l ent and or areta er r tten u ulat e ro ort on o ba el ne -day o n on ent BA B ro e onal and th al o l - a era e han e o nt dete t on and on den e an e ode or Beha or Analy t 201. nter al. hey a ly the e trate e n th art u- e r t on alone thou h nu ent. lar tudy to loo n at te - ount data h le h h- o on n e our el e and our l entele that the l ht n the ad anta e and d ad anta e o ea h le ented nter ent on a re on ble or the strategy. alterat on n the tar eted beha or e u t ta e e el trend and ar ab l ty are urther llu - another step—the step inherent in the term analy- trated n the e t on n ha ter 9 on e on trat n n a l ed beha or analy by e loy n un t onal Relat on and General Re u re ent or an e er ental de n. e learn n hat e er - S n le-Sub e t e n. ental de n are and ho they do the r ob are e ent al ll or e ery beha or analy t. hat the ub e t o our ne t ha ter. SUMMARY AND CONCLUSIONS nder tand n and nter ret n ra h and a ro- r ately re ord n ra h n and nter ret n data